a study of the effect of blasting vibration on green concrete

A STUDY OF THE EFFECTSOF BLASTING VIBRATION

ON GREEN CONCRETE

GEO REPORT No. 102

A.K.H. Kwan & P.K.K. Lee

GEOTECHNICAL ENGINEERING OFFICE

CIVIL ENGINEERING DEPARTMENT

THE GOVERNMENT OF THE HONG KONG

SPECIAL ADMINISTRATIVE REGION

A STUDY OF THE EFFECTSOF BLASTING VIBRATION

ON GREEN CONCRETE

GEO REPORT No. 102

A.K.H. Kwan & P.K.K. Lee

This report was prepared by A.K.H. Kwan and P.K.K. Lee inSeptember 1998 under Consultancy Agreement No. CE 21/94

for the sole and specific use of the Government of theHong Kong Special Administrative Region

- 2 -

© The Government of the Hong Kong Special Administrative Region

First published, April 2000

Prepared by:

Geotechnical Engineering Office,Civil Engineering Department,Civil Engineering Building,101 Princess Margaret Road,Homantin, Kowloon,Hong Kong.

This publication is available from:

Government Publications Centre,Ground Floor, Low Block,Queensway Government Offices,66 Queensway,Hong Kong.

Overseas orders should be placed with:

Publications Sales Section,Information Services Department,Room 402, 4th Floor, Murray Building,Garden Road, Central,Hong Kong.

Price in Hong Kong: HK$70Price overseas: US$12 (including surface postage)

An additional bank charge of HK$50 or US$6.50 is required per cheque made in currenciesother than Hong Kong dollars.

Cheques, bank drafts or money orders must be made payable toThe Government of the Hong Kong Special Administrative Region.

- 5 -

EXECUTIVE SUMMARY

In order to study the effects of blasting vibration on concrete including freshly placedconcrete (green concrete) so that blasting vibration control limits may be established to avoidcausing damage to concrete structures, a literature review and an experimental programmehave been conducted.

The findings of the literature review have been reported separately in the LiteratureReview Report submitted in October 1994. Basically, very large discrepancy exists in thevibration control limits currently adopted by different building authorities. There have beenrelatively few rigorous studies on the possible effects of blasting vibration on concrete and upto now the vibration resistance of concrete has never been directly measured. Nevertheless,putting the scanty results of previous studies together, it may be concluded that whenconsidering the effects of blasting vibration on concrete, concrete structures should beclassified into unrestrained structures (aboveground structures) which are free to vibrate andwould respond like being attacked by an earthquake and restrained structures (undergroundstructures) which are forced to deform with the surrounding soil. The response of anunrestrained structure depends on the structural form and thus dynamic analysis is required toevaluate the effects of blasting vibration on it. On the other hand, since restrained structureswould deform with the same strain as that of the ground, the effects of blasting vibration aredependent on the ground strain produced by the propagation of the shock wave through theground. Regarding the estimation of vibration resistance of concrete, most engineers assumethat the vibration resistance of concrete is proportional to the strength of the concrete but it isnot quite clear whether the strength should be compressive strength or tensile strength. Thesituation is rather confusing and to date no rational framework exists for establishingvibration control limits to avoid causing blasting damage to nearby concrete structures.

The experimental programme involved the testing of 6 typical concrete mixes and1440 concrete specimens. A method of testing concrete for their shock vibration resistanceby subjecting prisms cast of the concrete to hammer blows was developed. Vibrationintensities up to a peak particle velocity of 3000 mm/s have been applied. After hammering,the prismatic specimens were continued to be cured until the age of 28 days and then testedfor their compressive and tensile strengths. It was found that even at intensities high enoughto break the concrete specimen into pieces, the shock vibrations applied have little effect onthe compressive strength of concrete. Vibration damages were mainly in the form ofcracking and reduction in tensile strength. Hence, the vibration resistance of concrete isdependent more on tensile strength than compressive strength. The vibration resistanceresults (in terms of peak particle velocity of vibration) were quite scattered but the largenumber of experimental data allowed us to determine statistically the relationship between thelower bounds or characteristic values of the shock vibration resistance to concrete age,compressive strength, tensile strength and ultrasonic wave velocity with fairly goodcorrelation coefficients. This is the first time that the vibration resistance of concrete iscorrelated to other mechanical properties. Comparison of the vibration resistance values ofdifferent types of concrete with or without PFA and cured at different temperatures revealedthat the correlation of vibration resistance to other mechanical properties is not dependent onPFA content or curing temperature. Thus, a single set of vibration control criterion may beapplied to concretes within the range studied herein. From these vibration resistance results,vibration control limits that may be applied to Hong Kong and perhaps also to other places

- 6 -

may be obtained by dividing the vibration resistance values by a suitable factor of safety. Inview of the large variation of vibration resistance results, a factor of safety of 5 isrecommended. Despite the use of a fairly large factor of safety, the vibration control limitsso derived are higher than most existing limits currently being used in actual construction.Thus if the vibration control limits derived in the present study are adopted, higher intensityblasting and therefore quicker and more economical construction would be allowed.

It should, however, be borne in mind that so far no field testing has been carried out.Before the recommended vibration control limits are applied, it should be prudent to firstconduct field tests to confirm the safety of their use. Soil type and possible soil-structureinteraction should be included in further studies.

- 7 -

CONTENTS

PageNo.

Title Page 1

PREFACE 3

FOREWORD 4

EXECUTIVE SUMMARY 5

CONTENTS 7

1. INTRODUCTION 9

2. BACKGROUND OF STUDY AND RESEARCH SIGNIFICANCE 11

3. EXPERIMENTAL PROGRAMME 15

3.1 Concrete Mixes Tested 15

3.2 Tests Carried Out 16

3.2.1 Details of the Hammering Tests 19

3.2.2 Details of the Stiffening Time Tests 23

3.2.3 Details of the Integrity Tests 23

3.2.4 Test Standards Followed 24

4. TEST RESULTS 26

4.1 Quality Control Tests of Concrete Mixes Cast 26

4.2 Stiffening Times of the Concrete Mixes 33

4.3 Material Properties at Time of Hammering Test 42

4.4 Results of Hammering Tests and Integrity Tests 61

5. VIBRATION RESISTANCE OF THE CONCRETE MIXES TESTED 87

5.1 Effects of Shock Vibration on Concrete 87

5.2 Lower Bound Estimates of Vibration Resistance 115

5.3 Recommended Vibration Control Limits 118

6. THEORETICAL ANALYSIS 125

6.1 Mean Value Estimates of Vibration Resistance 125

6.2 Vibration Resistance of Concrete at Age ≤ 12 Hours 132

- 8 -

PageNo.

6.3 Vibration Resistance of Concrete at Age ≥ 12 Hours 135

7. DISCUSSIONS 146

7.1 Failure Criterion for Vibration Resistance 146

7.2 Vibration Resistance During First Few Hours 147

7.3 Material Parameters Determining Vibration Resistance 148

7.4 Restrained and Unrestrained Structures 149

8. CONCLUSIONS AND RECOMMENDATIONS FOR FURTHER STUDY 152

8.1 Conclusions 152

8.2 Recommendations for Further Study 158

9. REFERENCES 159

- 9 -

1. INTRODUCTION

This research study was carried out by The University of Hong Kong for GeotechnicalEngineering Office, Civil Engineering Department of the Hong Kong Government. It wasinitiated by Special Projects Division of the Geotechnical Engineering Office and waslaunched in June 1994 after Agreement No. CE21/94 was signed between The University ofHong Kong and the Hong Kong Government.

The scope of the study was specified in the Brief of Agreement No. CE21/94.According to Clause 2 of the Brief, the aim of the study is to develop a rational framework ormethodology for assessing the effects of blasting on green concrete and for setting upappropriate control criteria in Hong Kong. The study consists mainly of the following threeparts:

(1) a literature review covering the presently available data, existing methods of assessingthe effects of vibration on green concrete, and vibration control criteria being adoptedin Hong Kong and elsewhere;

(2) an experimental programme of subjecting typical concrete used in Hong Kong to highintensity vibrations to evaluate their vibration resistances at different ages andinvestigate the short/long term effects of vibration on concrete; and

(3) a theoretical study to correlate the vibration resistance of a concrete to its mechanicalproperties and develop a method of setting up appropriate vibration control criteriabased on the known properties of concrete.

As required by Clauses 5.1 and 5.2 of the Brief, an Inception Report on the proposedmethod of study, detailed testing programme and schedule of works had been submitted inJuly 1994. Subsequently, the study was carried out to completion following generally themethod outlined in the Inception Report except for some minor details which are explained inrelevant parts of this report.

As originally scheduled in the Inception Report, the study first started with theliterature review. After three months of study, a 165-page Literature Review Report No.1was completed and submitted in accordance with Clause 5.3 of the Brief in October 1994.Although the literature review was substantially completed after the submission of the report,it is being continued till the end of the Assignment so that the most updated information onthe topic can be included. Further findings of the literature review will be reported in theLiterature Review Report No. 2 which will be submitted separately.

Preparation for the laboratory tests started in August 1994. A number of testingequipment, including 15 steel moulds for concrete prisms, a specially designed hammeringtester with an automatic hammer release device, a high-g (high acceleration range)accelerometer together with its conditioning amplifier, an A/D conversion card for vibrationdata acquisition, an electronic stiffening time tester, a direct tension testing equipment and a486-PC computer for controlling the hammer release operation and the data acquisitionprocess, were purchased/fabricated. By January 1995, all the necessary equipment wereacquired and ready for use.

However, due to the lack of a suitable research assistant to help with the laboratoryworks, the tests did not start until June 1995. Two Research Assistants, Mr. G.L. Ni and

- 10 -

Mr. W. Zheng, both from Guangdong Provincial Research Institute of Water Conservancyand Hydro-power, had been employed to perform the actual testing. Together, they havecontributed twelve man-month’s of hard work towards the successful completion of theexperimental programme. In total, 720 concrete prisms, 432 concrete cubes and 288concrete cylinders have been cast and tested. All the testing works were carried out in HongKong using local materials under the supervision of Dr. A.K.H. Kwan and Mr. P.K.K. Lee,both Senior Lecturers of Department of Civil and Structural Engineering, The University ofHong Kong. Details of the experimental programme, the raw and processed data, theoreticalstudies carried out, the findings arrived at and the recommendations made from the work areall presented in this Final Report, which is written jointly by Dr. A.K.H. Kwan andMr. P.K.K. Lee. All the data processing, graph plotting and drawings in the report weredone by Mr. W. Zheng, who has spent three whole months to help completing this report.

- 11 -

2. BACKGROUND OF STUDY AND RESEARCH SIGNIFICANCE

With the fast-track construction in Hong Kong, programme and operation constraintsoften call for blasting to be carried out in close proximity to freshly cast concrete piles andstructures. Ground vibration caused by blasting may induce cracks and may have otheradverse effects on the green concrete. This has raised concern among the constructionindustry the possible adverse effects of blasting vibration on the integrity of green concrete.

The construction industry in Hong Kong tends to adopt very restrictive limits on themaximum ground vibration induced by blasting on a freshly cast concrete structure. Thelimits are generally based on experience and limited available literature from overseascountries. However, site conditions in those countries may not be strictly applicable locallyin Hong Kong. Therefore, there is a need to carry out a comprehensive research study toassess the effects of blasting vibration on green concrete.

"Green concrete", which may be taken to mean freshly placed concrete or morespecifically concrete having an age of less than 24 hours (Hulshizer and Desai 1984), isparticularly vulnerable to weakening of its properties if subjected to high intensity shockvibration that could disrupt the concrete matrix during the formative bond developmentprocess. The possible effects of shock vibration on curing concrete, especially freshlyplaced concrete, are very complicated and up to now, little is known about these effects.This is reflected in the large differences in the permitted peak particle velocity specified bydifferent building authorities.

Nevertheless, most engineers accept that fully hardened concrete can withstand shockvibration up to a peak particle velocity of around 100 mm/s without detrimental effects (Wiss1981). It is also often assumed that the resistance of "curing concrete" (concrete that has notyet fully developed its strength) to shock vibration is proportional to its strength and thus thepermitted peak particle velocity may be taken as proportional to the strength of the concrete.Based on this principle, Gamble and Simpson (1985), Bryson and Cooley (1985), andOlofesson (1988) established their criteria for permitted peak particle velocity as in Table 2.1.

- 12 -

Table 2.1 - Permitted peak particle velocity commonly adopted for concrete with agesof more than 12 hours

ReferencesPermitted peak particle velocity

at different ages (mm/s)

12hrs

1day

2days

3days

7days

28days

90days

Gamble andSimpson, 1985 5 10 20 30 50 100 100

Bryson andCooley, 1985 5 50 50 50 100 100 100

Olofesson, 1988- curing at 5ºC - - 8 11 35 80 100

Olofesson, 1988- curing at 21ºC - 14 30 40 60 85 100

Table 2.2 - Permitted peak particle velocity commonly adopted for green concrete(concrete with ages of less than 24 hours)

ReferencesPermitted peak particle velocity

at different times after placement (mm/s)

0-2hrs

4hrs

8hrs

12hrs

18hrs

24hrs

>24hrs

Gamble andSimpson, 1985 - 2 3 5 8 10 10

Bryson andCooley, 1985 5 5 5 5 5 5 50

Olofesson, 1988- curing at 5ºC 100 (0-10 hrs) No blasting within 30 m (10-70 hrs)

Olofesson, 1988- curing at 21ºC 100 (0-5 hrs) No blasting within 30 m (5-24 hrs) 14

Since the strength of concrete is relatively low during the first 24 hours after casting,blasting is usually greatly restricted near green concrete. On the other hand, however, it hasalso been suggested that within the first few hours before initial set, the freshly placedconcrete should be able to withstand shock vibration up to 100 mm/s (Olofesson 1988). Thepermitted peak particle velocity commonly adopted for green concrete are shown in Table 2.2.From the tabulated values, it can be seen that very large discrepancies exist between thevarious recommended vibration limits, especially those being applied to concrete cast within afew hours.

- 13 -

As described by Dowding (1985), an interesting experiment had been carried out byEsteves in 1978. The experiment revealed the surprising result that although there is aperiod of greater susceptibility to vibration cracking between 10 to 20 hours after placementof the concrete, the threshold vibration for cracks to appear during this period is as high as150 mm/s. Hulshizer and Desai (1984) had also carried out a testing program, whichincluded both laboratory and field tests, to evaluate the effects of high intensity shockvibrations on green concrete. It was found that although many concrete specimens weresubjected to shock vibrations up to and in the range of 200 to 250 mm/s, there was noevidence to indicate that the vibrated concrete would not perform like un-vibrated concrete.Accordingly, they suggested fairly high vibration limits of 100 mm/s, 38 mm/s, 50 mm/s, 100mm/s and 175 mm/s for concrete cast in less than 3 hours, within 3 - 11 hours, within 11 - 24hours, within 1 - 2 days and over 2 days respectively. Thus, on one hand, the first 24 hoursafter concrete placement is generally regarded as the most critical period and engineers areextremely cautious in controlling nearby blasting during this period, on the other hand, testresults have indicated that the green concrete can, in fact, withstand fairly high intensity shockvibration during the first few hours after casting. More research studies are needed in orderto clarify the situation.

Relatively speaking, the case for concrete placed more than 24 hours is more straightforward. However, it is not without problems. Firstly, although the assumption that the peakparticle velocity the curing concrete can withstand is proportional to the strength of theconcrete is quite commonly adopted, it has never been testified by any experimental results.It is also not clear whether the strength should be the compressive strength or the tensilestrength of the concrete. Since the most likely damage that will be caused by the shockvibration is cracking, the tensile strength should be more relevant as a measure of vibrationresistance. However, it seems that most engineers are still using the compressive strengthfor specifying the vibration limits. Secondly, there is also the problem that the maximumstress induced by the shock vibration is not really proportional to the peak particle velocity.According to Dowding (1984), the maximum dynamic strain εd induced by the shockvibration is given by:

εdvc

= ...(2.1)

where v is the peak particle velocity and c is the wave propagation velocity. Multiplying bythe dynamic modulus Ed , the maximum stress induced by the shock vibration is obtained as:

σ = Evcd

...(2.2)

Vibration cracking occurs when the maximum stress exceeds the dynamic tensile strength f tdof the concrete, i.e.

Evc

fd td> ...(2.3)

From this inequality, the threshold peak particle velocity v' for vibration cracking to occurcan be evaluated as:

vcEf

dtd' = ...(2.4)

- 14 -

Hence, the peak particle velocity that the curing concrete can withstand is not reallyjust a function of the tensile strength of the concrete. In determining the shock vibrationresistance of the concrete, it is necessary to take into account also the wave propagationvelocity and dynamic modulus of the concrete. As a matter of course, tests are required toverify the above arguments. Finally, the rate of strength development is dependent on manyfactors, e.g. mix composition and curing temperature, and thus it would be too simplistic toestablish just a single set of vibration limits regardless of the mix composition and curingtemperature. As concrete mixes containing PFA (pulverized fuel ash) may develop theirstrength at much slower rates than those without, it is necessary at least to distinguish betweenPFA concrete and non-PFA concrete. Tests on different types and grades of concrete undera curing temperature representative of the Hong Kong conditions are necessary so that theactual rates of strength development of the different concrete can be taken into account inestablishing the vibration limits.

In this study, a series of tests on concrete mixes of different grades, with or withoutPFA and at different ages has been carried out to determine their shock vibration resistance.Although the Brief asked for study on only green concrete (concrete having an age of lessthan 24 hours), concrete with ages of more than 24 hours but less than 28 days were alsoincluded in the study because it was felt that concrete at such ages might also be affected byvibration. The effects of the shock vibration applied to the concrete were determined bycontinuing to cure the concrete after they were subjected to the shock vibration until the ageof 28 days and testing the concrete for their tensile and compressive strengths. In order tocorrelate the vibration resistance of the concrete to their mechanical properties, they were alsotested for their mechanical properties if possible at the time of vibration test. As thevibration resistance of a green concrete was expected to be quite sensitive to its rate ofstiffening, stiffening time tests of the concrete were also carried out. After all the tests werecompleted, the vibration resistances of the various grades and types of concrete at differentages were evaluated and then correlated to their respective mechanical properties andstiffening times to see whether a theory for estimating the vibration resistance of concretecould be developed.

- 15 -

3. EXPERIMENTAL PROGRAMME

3.1 Concrete Mixes Tested

The concrete mixes tested are presented in Table 3.1 while the mix proportions of theconcrete mixes derived by trial mixing are listed in Table 3.2. They ranged from Grades 20to 40, contained either no or 25% PFA, but were all designed to have a slump of 75 mm. Allthe concrete mixes were made using local materials. The same sources of materials wereused throughout the study so as to maintain consistency.

Table 3.1 - Concrete mixes tested

Mix no. Grade Slump (mm) PFA content

G20/0PFA

G20/25PFA

G30/0PFA

G30/25PFA

G40/0PFA

G40/25PFA

20

20

30

30

40

40

75

75

75

75

75

75

0%

25%

0%

25%

0%

25%

Table 3.2 - Mix proportions of the concrete

Mix no. Water/binderratio

OPC content(kg/cu.m.)

PFA content(kg/cu.m.)

Fine to totalaggregate ratio

Superplasticizer(litre/cu.m.)

G20/0PFA

G20/25PFA

G30/0PFA

G30/25PFA

G40/0PFA

G40/25PFA

0.65

0.62

0.55

0.53

0.45

0.43

308

242

364

283

444

349

-

81

-

94

-

116

0.40

0.40

0.35

0.35

0.30

0.30

0.74

0.50

1.10

0.92

1.50

1.10

In order to study the effects of the curing temperature, two batches of concrete, onecured at a designated curing temperature of 20ºC and the other at a designated curingtemperature of 30ºC, were produced from each concrete mix. The two batches of concrete,though cast of the same mix proportion, are treated as different concretes and assigneddifferent batch numbers as shown in Table 3.3.

- 16 -

Table 3.3 - Concrete batch numbers

Batch no. Mix no. Curing temperature

G20/0PFA/20C

G20/0PFA/30C

G20/25PFA/20C

G20/25PFA/30C

G30/0PFA/20C

G30/0PFA/30C

G30/25PFA/20C

G30/25PFA/30C

G40/0PFA/20C

G40/0PFA/30C

G40/25PFA/20C

G40/25PFA/30C

G20/0PFA

G20/0PFA

G20/25PFA

G20/25PFA

G30/0PFA

G30/0PFA

G30/25PFA

G30/25PFA

G40/0PFA

G40/0PFA

G40/25PFA

G40/25PFA

20ºC

30ºC

20ºC

30ºC

20ºC

30ºC

20ºC

30ºC

20ºC

30ºC

20ºC

30ºC

For the sake of quality control, samples were taken from each batch of concrete forworkability tests and cube strength tests. The workability tests included slump test,compacting factor test and V-B time test. For the cube strength tests, three cubes were cast.The cubes were cured under a standard temperature of 27ºC (note that this curing temperatureis not the same as the designated curing temperature for the other test specimens which werecured at either 20ºC or 30ºC) as required by the Hong Kong Construction Standard CS1:1990,and tested at the age of 28-day.

3.2 Tests Carried Out

For each batch of concrete, the following laboratory tests have been carried out:

(1) Hammering tests - Prisms cast of the concrete mix and cured under the designatedcuring temperature were subjected to hammer blows of different intensity, and thepeak particle velocity required to fracture the concrete at various times after castingwere measured.

(2) Stiffening time tests - The penetration resistance of the mortar fraction sieved from theconcrete mix at various times after mixing were measured and hence the stiffeningtimes of the concrete mix determined so that the change in vibration resistance of theconcrete during the early ages could be related to the stiffening time.

(3) Material property tests - The compressive strength, tensile strength, dynamic modulus,static modulus and longitudinal wave propagation velocity of the concrete at the time

- 17 -

of performing hammering tests were measured if the concrete had hardenedsufficiently for the moulds to be removed (normally, the moulds could be removed at12 hours after casting) so that the vibration resistance of the concrete might be relatedto these parameters.

(4) Integrity tests - The concrete prisms which had been subjected to different intensity ofhammer blows were tested for their 28-day compressive and tensile strengths so as toevaluate the long term effects of the shock vibration on the integrity of the concrete.

The hammering tests on the concrete were carried out at the following ages:

2 hrs., 4 hrs., 6 hrs., 8 hrs., 10 hrs., 12 hrs.,18 hrs., 24 hrs., 3 days, 7 days, 28 days.

For concrete more than 12 hours old, the moulds of the concrete prisms were first removedbefore performing any tests. After the moulds were removed, ultrasonic tests were carriedout on the concrete prisms to measure the longitudinal wave propagation velocity anddynamic modulus of the concrete. Then the hammering tests were performed by subjectingthe concrete prisms (without their moulds) to hammer blows. At the time of performing thehammering tests, the cubes and cylinders cast of the same concrete and cured under the sameconditions were also tested to determine the compressive and tensile strengths of the concrete.However, since it was difficult to de-mould concrete specimens within 12 hours after casting,freshly placed concrete with ages of less than 12 hours were tested with their moulds on.Thus hammering tests on concrete placed within 12 hours were carried out by keeping theconcrete in their moulds and subjecting the concrete together with their moulds to the hammerblows. As the freshly placed concrete were still colloidal, their stiffness or strength were notmeasured. Instead, the stiffening time of the concrete was measured so that the change invibration resistance of the concrete as the concrete hardens could be related to the time whenthe concrete changed its state.

In order to deal with the situation that concrete with ages of less than 12 hours andthose with ages of more than 12 hours had to be treated differently, two separate series of testswere carried out on each batch of concrete. The first series of test was on concrete less than12 hours old, while the second series was on concrete more than 12 hours old. The testscarried out in each series of test are listed below:

Series 1:Quality control tests;Hammering tests (with moulds on) at:

2 hrs., 4 hrs., 6 hrs., 8 hrs., 10 hrs., 12 hrs.;Stiffening time tests on the fresh concrete;Integrity tests at 28-day.

Series 2:Quality control tests;Hammering tests (with moulds removed) at:

12 hrs., 18 hrs., 24 hrs., 3 days, 7 days, 28 days;Material property tests at time of performing hammering tests;Integrity tests at 28-day.

- 18 -

Basically, for each age at which hammering tests were to be carried out, five prismswere cast and cured at their designated curing temperature of either 20ºC or 30ºC until thetime for the hammering tests. One of the five prisms was treated as a control specimen thatwould not be subjected to any hammer blow but would be cured alongside the others forfuture comparison testing. Among the other four prisms, one was subjected to a number ofhammer blows with increasing intensity until the first crack appeared in order to measure thevibration intensity at which the concrete cracked and study the cumulative effects ofhammering vibrations on the concrete. The remaining three prisms were each subjected toone hammer blow of different intensity to measure the peak particle velocity required to causecracking of the concrete. As far as possible, the hammer blows were applied to the threeprisms in such a way that at least one but not more than two of the prisms would be cracked.However, due to the erratic nature of vibration resistance, quite often none of the three prismswhich had each been subjected to one hammer blow cracked. In order to produce crackingso that the vibration intensity necessary to cause cracking may be determined, some of theprisms were re-hammered (i.e. subjected to more than one hammer blows) until at least one ofthem cracked. After the hammering tests, the four prisms which had been subjected to thehammering were continued to be cured together with the control specimen at their designatedcuring temperature. At the age of 28-day, all the prisms were tested for their tensile andcompressive strengths so as to evaluate the long term effects of the hammer blows on theintegrity of the concrete.

In addition to the above, for Series 1 tests, one fresh concrete sample was taken fromeach batch of concrete for stiffening time test. The stiffening time test was carried out on themortar fraction of the concrete which was separated from the fresh concrete specimen bysieving and then kept at the same temperature as the designated curing temperature of theconcrete (i.e. either 20ºC or 30ºC) at all times until the test was finished. The penetrationresistance of the mortar was measured using a stiffening time tester at half-hourly intervalsuntil the mortar has completely stiffened. The respective stiffening times to reachpenetration resistances of 0.5 and 3.5 MPa were determined by plotting the penetrationresistance against time and reading the time values corresponding to 0.5 and 3.5 MPapenetration resistance from the graph.

Apart from the prisms required for hammering tests, in the Series 2 tests, three cubesand three cylinders were also cast for each age at which the hammering tests were carried out.The cubes and cylinders were placed alongside the prisms during curing so that they werecured under exactly the same conditions as for the prisms. At the time of performing thehammering tests on the prisms, the three cubes were crushed to determine their compressivestrength and the three cylinders split to determine their split cylinder tensile strength.Ultrasonic tests were also carried out on the four concrete prisms prior to applying thehammer blows to them to measure the longitudinal wave propagation velocity and dynamicmodulus of the concrete. In addition, for measurement of static modulus, one extra cylinderwas cast from each batch of concrete. It was cured under the same condition as for the otherspecimens and its static modulus measured at the time of performing hammering tests.

More details of the above tests are given in the following sub-sections.

- 19 -

3.2.1 Details of the Hammering Tests

There is no standard test method for measuring the vibration resistance of concrete.Among the existing methods, the simplest and yet most direct way of measuring the vibrationresistance of concrete is to subject concrete prisms to hammer blows in the longitudinaldirection to induce longitudinal wave stresses along the prisms and then observe the formationof cracks on the concrete surfaces. This hammering test method, with a set up as shown inFig.3.1, was adopted in this study.

The dimensions of the prisms were 150 mm × 150 mm × 750 mm long which arethe same as those of a standard test beam specified in BS1881: Part 118: 1983 for flexural testand equivalent cube test. The reasons for adopting such standard beam size were that thisallowed the standard beam moulds to be used in the present study and the prisms to be testedfor their integrity in accordance with BS1881: Part 118.

The hammer was mounted on a swing arm for hitting one end of the test prismperpendicularly. Its weight was adjustable between a minimum weight of 7 kg to a

Figure 3.1 - Schematic of set up for hammering test

maximum weight of 36 kg and its height when it was released was also adjustable from 200mm to 800 mm above the striking point. Hence by varying the weight and/or the height ofthe hammer when released, the impact energy to be applied to the test prism could be adjusted

- 20 -

between 14 to 280 Joules. To ensure that the hammer blow impact was applied uniformly tothe cross-section of the test prism, a steel plate of 10 mm thick was attached to the end of theprism to be struck. Concrete cast within 12 hours were tested with their moulds on whileconcrete cast more than 12 hours were tested with their moulds removed. The concreteprisms were mounted on roller bearings so that they were free to move longitudinally whensubjected to hammer blows. Soft rubber pads were fixed at the far end of the testing frameto restrain the prisms from falling off the testing frame after being subjected to hammerblows. The whole hammering operation was controlled by a 486-PC computer which wasalso used for data acquisition. Before the hammering operation, the hammer was liftedmanually and hooked to an automatic hammer release device. At the trigger of an electronicsignal from the controlling computer, the hammer was released from the hook of theautomatic hammer release device and the data acquisition process started. This ensured thatthe hammering operation and the data acquisition process were synchronized. Details of thehammer release device are shown in Fig.3.2.

An accelerometer was attached to the test prism to measure the intensity and frequencyspectrum of the shock vibration induced by the hammer blow. To fix the accelerometer tothe test prism, a perspex seating was first glued onto the top surface of the prism and then theaccelerometer attached to the perspex by a small screw. Three different locations for fixingthe accelerometer, at the striking end, at mid-length and at the far end, had been tried but itturned out that the mid-length location gave the best results. Added with the considerationthat any cracks formed were usually near the centre of the prisms, it was decided to attach theaccelerometer at mid-length of the prism, as shown in Fig.3.3. After the application of eachhammer blow to the test prism, the prism was inspected for any transverse cracks formed onthe concrete.

Figure 3.2 - The hammer release device

- 21 -

Figure 3.3 - Attached position of the accelerometer

All prisms were continued to be cured under their respective designated curingtemperatures after the hammering tests. In order to make sure that any cracks formed byshock vibration during the hammering tests were closed so as to allow for any possible self-healing effects, the prisms were cured with their longitudinal axis oriented vertically so thatthe cracks were kept closed under the self weight of the prisms. At the age of 28-day, theprisms were each tested for their compressive and tensile strengths to evaluate the long termeffects of the shock vibration on the integrity of the concrete (for details of the integrity tests,refer to Section 3.2.3).

The accelerometer used was a Bruel & Kjaer type 8309s piezoelectric high frequencyresponse, high-g (100,000g) shock vibration accelerometer. It was connected to a Bruel &Kjaer type 2626 signal conditioning amplifier which drove the accelerometer and gaveanalogue output for the acceleration measurement. The analogue output from theconditioning amplifier was fed into a 12-bit A/D converter installed inside the controllingcomputer which converted the acceleration signal into digital form at a rate of 100 kHz. Thedigital signals from the A/D converter were first stored into a hard disk and then processedusing a software called GLOBAL LAB. In order to obtain the peak particle velocity of theshock vibration, the acceleration results were numerically integrated with respect to time toyield the particle velocity at the point of measurement. The particle velocity so derivedconsisted of two components: the velocity of the rigid body movement of the test prism andthe vibration velocity of the longitudinal shock wave, as illustrated in Fig.3.4. As the rigidbody movement of the prism would not induce any stress, it had no effect on the concrete andshould thus be removed. A trend removal process was used to remove the rigid bodymovement component (the trend of the particle velocity results was actually the rigid bodymovement of the prism) from the particle velocity results so as to yield the pure wave velocitycomponent. It was this pure wave velocity component that induced cyclic tensile-compressive stresses in the concrete prism and caused cracking or damage of the concrete.The pure wave velocity component so obtained generally showed the pattern of a typicalattenuating wave with a peak amplitude occurring within the first one or two cycles ofvibration and a gradually decreasing wave amplitude with time. From the pure wavevelocity component, the peak particle velocity of the shock wave was determined and taken asa measure of the intensity of the applied shock vibration.

- 22 -

Acceleration

-5000

-4000

-3000

-2000

-1000

0

1000

2000

3000

4000

5000

0 2 4 6 8 10 12

Time (ms)

Acc

eler

atio

n (m

/s2 )

Particle velocity

-500

0

500

1000

1500

2000

2500

3000

0 2 4 6 8 10 12

Time (ms)

Parti

cle

velo

city

(mm

/s)

Wave component of particle velocity

-800

-600

-400

-200

0

200

400

600

800

1000

0 2 4 6 8 10 12

Time (ms)

Wav

e co

mpo

nent

of p

artic

le v

eloc

ity

(mm

/s)

Figure 3.4 - Obtaining particle velocity of shock wave from acceleration result

- 23 -

3.2.2 Details of the Stiffening Time Tests

The stiffening time tests were carried out in accordance with Appendix C of BS5075:Part 1: 1982. They were for the determination of the time after mixing of the concrete forthe mortar fraction sieved from the concrete mix to reach penetration resistances of 0.5 MPaand 3.5 MPa. According to BS5075, the time for the mortar fraction to reach a penetrationresistance of 0.5 MPa corresponds approximately to the extreme limit for placing andcompacting concrete, while the time to reach a penetration resistance of 3.5 MPa gives aguide to the time available for the avoidance of cold joints. Although the stiffening timemeasurements are originally intended to be used for scheduling of concreting operations, it isfelt that they might also be useful for setting vibration limits at various times after placing ofthe concrete.

As the curing temperature could significantly affect the stiffening time of the concrete,the mortar samples were kept at all times under the designated curing temperature of either20ºC or 30ºC, i.e., the same curing temperature as for the concrete prisms to be subjected tohammering tests.

The stiffening time tester used consisted of a rig for lowering a pin of dimensionspecified by the British Standard vertically into a can of mortar and an electronic balance ofcapacity 50 kg for measuring the penetration resistance of the mortar.

3.2.3 Details of the Integrity Tests

After the hammering tests, all prisms, including those which had been subjected tohammer blows and the control specimens which had not been subjected to any vibration, werecontinue to be cured at their designated curing temperature until the age of 28-day when theywould be tested for their tensile and compressive strengths so as to evaluate the long termeffects of the hammer blows on the integrity of the concrete. Theoretically, the tensile andcompressive strengths of the concrete prisms could be determined by the flexural test methodspecified in BS1881: Part 118: 1983 and the equivalent cube test method specified in BS1881:Part 119: 1983. However, since the concrete prisms might have been cracked during thehammering tests and the exact positions of any cracks formed, which could be anywherealong the length, might affect the flexural strength measurements, it was likely that the tensilestrength results obtained by the flexural test method would be quite erratic. To avoid thisproblem, the tensile strength of the prisms was measured by applying direct tension to theends of the prisms so that the tensile strength results would be independent of the positions ofany cracks present in the specimens. After the direct tension test, the equivalent cubemethod was applied to portions of the prisms broken in tension to determine the equivalentcube strength of the prisms.

- 24 -

Figure 3.5 - Direct tension testing equipment

A pair of grips has been specifically fabricated for the direct tension test. Each gripconsisted of two steel plates and two rubber sheets to be attached to two opposite faces at oneend of the prism. Gripping force was provided at each grip by six bolts tightened to pressthe steel plates and rubber sheets against the prism, as shown in Fig.3.5. A hook whosecentre coincided with the centroidal axis of the test prism was provided at each grip to makesure that the force applied would not cause bending of the test specimen. The grips, togetherwith the test prism that they were holding, were installed onto an Avery 1000 kN capacityuniversal testing machine and then tension load was applied to the specimen under loadcontrol until the prism failed. None of the specimens failed at the gripping positionsindicating that the gripping stresses applied at the two ends of the prisms had not affected thetension failure of the specimens.

3.2.4 Test Standards Followed

As far as possible, relevant British or local standards were followed in carrying out thetests. The standards followed are listed below:

- 25 -

Quality control tests:

In general accordance with the following standards:

slump test BS1881: Part 102: 1983compacting factor test BS1881: Part 103: 1983V-B time test BS1881: Part 104: 1983cube test BS1881: Part 116: 1983

except that the curing temperature for the cubes was 27ºC, as required by the Hong KongConstruction Standard CS1: 1990.

Stiffening time test:

In general accordance with Appendix C of BS5075: Part 1: 1982 except that the storagetemperature of the mortar specimens for stiffening time tests was the same as the designatedcuring temperature of the concrete (the standard requires a storage temperature of 20ºC) andthat the penetration resistance measurements was performed at half-hourly intervals until 6hours after mixing of the concrete.

Material property tests:

In general accordance with the following standards:

ultrasonic test BS1881: Part 203: 1986cube test BS1881: Part 116: 1983split cylinder test BS1881: Part 117: 1983

except that the curing temperature of the test specimens was the same as the designated curingtemperature of the concrete.

Hammering test:

There was no relevant standard that could be followed.

Integrity tests:

The size of the prism specimens was the same as that of the standard 150 mm × 150 mm ×750 mm test beam specified in BS1881: Part 118: 1983. Casting of the prism specimens wasin accordance with BS1881: Part 109: 1983. The tensile strength of the prisms was,however, measured by applying direct tension instead of by the flexural testing methodspecified in BS1881: Part 118: 1983. Equivalent cube strength of the prisms was measuredin accordance with BS1881: Part 119: 1983. The curing temperature of the prisms was thedesignated curing temperature of the concrete instead of 20ºC required by the Britishstandards.

- 26 -

4. TEST RESULTS

4.1 Quality Control Tests of Concrete Mixes Cast

From each batch of concrete cast, samples were taken for workability and cubestrength tests for quality control purpose.

Three workability tests, namely: slump test, V-B time test and compacting factor test,were carried out. Although the concrete mixes were designed to have a slump of 75 mm, theactual slump values varied from about 65 mm to 85 mm. The corresponding V-B times andcompacting factors varied from about 1.0 s to 3.0 s and from about 0.92 to 0.97 respectively.These are considered reasonable variations for the large number of batches of differentconcrete mixes cast.

Three 150 mm cubes were cast from each batch of concrete for the strength test. Thecubes were de-moulded one day after casting and were then cured at 27ºC in a water tankuntil testing at the age of 28-day. From each set of the three cubes, the cube strength resultswere averaged to give the mean cube strength value for the batch of concrete cast.

Details of the test results for each batch of concrete cast are tabulated in Tables 4.1 to4.6. When studying the test results, it should be borne in mind that from each batch ofconcrete, fifteen prisms were cast. The fifteen prisms were used for hammering tests at threedifferent ages (the number of prisms required for hammering test at each age was five) andthus each set of quality control tests corresponded to concrete cast for hammering tests atthree different ages. That is why in Tables 4.1 to 4.6, concretes cast for hammering tests atdifferent ages can have the same quality control test results. It should also be noted that thecubes cast for quality control testing purpose were cured at a temperature of 27ºC as requiredby Hong Kong Construction Standard CS1 which is not the same as the designated curingtemperature for the concrete prisms cast for hammering tests.

- 27 -

Table 4.1 - Quality control test results for G20/0PFA/20C and G20/0PFA/30C

Batch no.Age at which

hammering testcarried out

Slump(mm)

V-B time(s)

Compactingfactor

Mean 28-daycube strength

(MPa)

2 hrs 75 1.5 0.97 32.7

4 hrs 75 1.5 0.97 32.7

G20/0PFA/20C 6 hrs 75 1.5 0.97 32.7

Series 1 tests 8 hrs 70 2.8 0.96 31.1

10 hrs 70 2.8 0.96 31.1

12 hrs 70 2.8 0.96 31.1

12 hrs 70 1.9 0.97 32.5

18 hrs 70 1.9 0.97 32.5

G20/0PFA/20C 24 hrs 80 1.5 0.97 31.7

Series 2 tests 3 days 70 1.9 0.97 32.5

7 days 80 1.5 0.97 31.7

28 days 80 1.5 0.97 31.7

2 hrs 80 1.4 0.96 31.8

4 hrs 85 1.0 0.97 31.6

G20/0PFA/30C 6 hrs 85 1.0 0.97 31.6

Series 1 tests 8 hrs 80 1.4 0.96 31.8

10 hrs 70 2.0 0.96 31.3

12 hrs 65 2.2 0.95 32.2

12 hrs 65 2.2 0.95 32.2

18 hrs 80 1.4 0.96 31.8

G20/0PFA/30C 24 hrs 75 2.3 0.97 31.9


7 days 75 2.3 0.97 31.9

28 days 75 2.3 0.97 31.9

- 28 -




Slump(mm)

V-B time(s)

Compactingfactor


(MPa)

2 hrs 70 1.7 0.96 20.5

4 hrs 70 2.1 0.94 23.5

G20/25PFA/20C 6 hrs 70 2.1 0.94 23.5

Series 1 tests 8 hrs 75 1.6 0.96 21.9

10 hrs 75 1.9 0.96 22.8

12 hrs 75 1.9 0.96 22.8

12 hrs 75 1.9 0.96 22.8

18 hrs 70 2.1 0.94 23.5

G20/25PFA/20C 24 hrs 70 1.7 0.96 20.5


7 days 75 1.1 0.95 20.7

28 days 75 1.1 0.95 20.7

2 hrs 70 2.3 0.96 24.6

4 hrs 75 2.8 0.96 26.6

G20/25PFA/30C 6 hrs 75 2.8 0.96 26.6

Series 1 tests 8 hrs 70 2.3 0.96 24.6

10 hrs 80 1.5 0.96 28.0

12 hrs 80 1.5 0.96 28.0

12 hrs 80 1.5 0.96 28.0

18 hrs 75 2.8 0.96 26.6

G20/25PFA/30C 24 hrs 70 2.1 0.94 24.3


7 days 70 2.1 0.94 24.3

28 days 70 2.1 0.94 24.3

- 29 -




Slump(mm)

V-B time(s)

Compactingfactor


(MPa)

2 hrs 75 2.5 0.94 40.7

4 hrs 75 2.5 0.94 40.7

G30/0PFA/20C 6 hrs 70 2.6 0.95 41.6

Series 1 tests 8 hrs 70 2.6 0.95 41.6

10 hrs 75 2.1 0.95 43.4

12 hrs 75 2.1 0.95 43.4

12 hrs 75 2.1 0.95 43.4

18 hrs 75 2.5 0.94 40.7

G30/0PFA/20C 24 hrs 70 2.6 0.95 41.6


7 days 75 2.8 0.95 42.7

28 days 75 2.8 0.95 42.7

2 hrs 75 2.3 0.95 41.5

4 hrs 75 2.3 0.95 41.5

G30/0PFA/30C 6 hrs 75 1.8 0.96 40.1

Series 1 tests 8 hrs 75 1.8 0.96 40.1

10 hrs 70 2.6 0.96 43.8

12 hrs 70 2.6 0.96 43.8

12 hrs 70 2.6 0.96 43.8

18 hrs 75 2.3 0.95 41.5

G30/0PFA/30C 24 hrs 75 1.8 0.96 40.1


7 days 75 2.0 0.96 43.3

28 days 75 2.0 0.96 43.3

- 30 -




Slump(mm)

V-B time(s)

Compactingfactor


(MPa)

2 hrs 75 2.6 0.93 29.8

4 hrs 70 3.0 0.95 35.2

G30/25PFA/20C 6 hrs 70 3.0 0.95 35.2

Series 1 tests 8 hrs 70 3.0 0.95 35.2

10 hrs 75 2.6 0.93 29.8

12 hrs 75 2.6 0.93 29.8

12 hrs 75 2.3 0.95 35.9

18 hrs 70 3.2 0.95 37.5

G30/25PFA/20C 24 hrs 75 2.3 0.95 35.9


7 days 75 2.3 0.95 35.9

28 days 70 3.2 0.95 37.5

2 hrs 65 3.2 0.95 38.5

4 hrs 70 2.3 0.95 36.8

G30/25PFA/30C 6 hrs 70 2.3 0.95 36.8

Series 1 tests 8 hrs 70 2.3 0.95 36.8

10 hrs 65 3.2 0.95 38.5

12 hrs 65 3.2 0.95 38.5

12 hrs 70 2.3 0.95 38.8

18 hrs 70 2.3 0.95 38.8

G30/25PFA/30C 24 hrs 65 3.2 0.94 37.0


7 days 65 3.2 0.94 37.0

28 days 65 3.2 0.94 37.0

- 31 -




Slump(mm)

V-B time(s)

Compactingfactor


(MPa)

2 hrs 65 2.0 0.94 51.7

4 hrs 75 1.5 0.93 53.8

G40/0PFA/20C 6 hrs 75 2.0 0.92 47.6

Series 1 tests 8 hrs 65 2.5 0.92 55.7

10 hrs 75 2.0 0.92 47.6

12 hrs 75 1.5 0.93 53.8

12 hrs 65 2.5 0.92 55.7

18 hrs 65 2.0 0.94 51.7

G40/0PFA/20C 24 hrs 70 2.5 0.92 55.7


7 days 75 2.0 0.92 47.6

28 days 75 1.5 0.93 53.8

2 hrs 75 1.9 0.97 52.4

4 hrs 75 2.0 0.97 51.2

G40/0PFA/30C 6 hrs 75 1.9 0.93 51.7

Series 1 tests 8 hrs 70 2.9 0.97 50.1

10 hrs 75 1.9 0.93 51.7

12 hrs 75 2.0 0.97 51.2

12 hrs 70 2.9 0.97 50.1

18 hrs 75 1.9 0.97 52.4

G40/0PFA/30C 24 hrs 70 2.9 0.97 50.1


7 days 75 1.9 0.93 51.7

28 days 70 2.0 0.97 51.2

- 32 -




Slump(mm)

V-B time(s)

Compactingfactor


(MPa)

2 hrs 70 2.0 0.94 45.1

4 hrs 70 2.0 0.94 45.1

G40/25PFA/20C 6 hrs 75 2.9 0.95 43.8

Series 1 tests 8 hrs 65 3.3 0.90 44.1

10 hrs 75 1.9 0.95 45.2

12 hrs 75 1.9 0.95 45.2

12 hrs 75 1.9 0.95 45.2

18 hrs 65 3.3 0.90 44.1

G40/25PFA/20C 24 hrs 70 2.0 0.94 45.1


7 days 65 3.3 0.90 44.1

28 days 75 2.9 0.95 43.8

2 hrs 70 2.1 0.95 46.2

4 hrs 70 2.1 0.95 46.2

G40/25PFA/30C 6 hrs 70 1.9 0.95 47.5

Series 1 tests 8 hrs 70 1.9 0.95 47.5

10 hrs 75 1.8 0.95 48.9

12 hrs 75 1.8 0.95 48.9

12 hrs 75 1.8 0.95 48.9

18 hrs 75 2.0 0.95 45.3

G40/25PFA/30C 24 hrs 70 1.9 0.95 47.5


7 days 75 2.0 0.95 45.3

28 days 75 2.0 0.95 45.3

- 33 -

A summary of the quality control test results is presented in Table 4.7 wherein theaverage values, standard deviations and coefficients of variation of the 28-day cube strengthsof the concrete mixes studied are listed. It is seen that the largest standard deviation of thevarious concrete mixes was only 2.8 MPa. This indicates that the production of the concretemixes was in generally under good quality control. The concrete mixes without PFAgenerally have their respective mean strengths more than 10 MPa higher than their specifiedcharacteristic strengths. However, the concrete mixes with PFA added were of somewhatinferior quality; their mean strengths were generally less than 10 MPa higher than theirspecified characteristic strengths. Thus when interpreting the test results, reference shouldbe made to their actual mean strengths rather their specified grades.

Table 4.7 - Statistical analysis of the 28-day cube strength results

Mix no.Average value of

28-day cube strength(MPa)

Standard deviation of28-day cube strength

(MPa)

Coefficient ofvariation

(%)

G20/0PFA

G20/25PFA

G30/0PFA

G30/25PFA

G40/0PFA

G40/25PFA

31.9

23.7

42.1

36.2

51.8

45.8

0.6

2.5

1.4

2.8

2.4

1.7

1.9

10.5

3.3

7.7

4.6

3.7

4.2 Stiffening Times of the Concrete Mixes

The stiffening time tests were conducted on the mortar fractions of the concrete mixeswhich were obtained by sieving the coarse aggregate particles away from the fresh concretes.Although the British Standard requires only the times to reach penetration resistances of 0.5 MPa and 3.5 MPa to be measured, the penetration resistance of the mortar was measuredat 30 minute intervals to see how the penetration resistance actually varied with time. Theresults of such measurements are plotted in Figs. 4.1 to 4.6. It can be seen from these figuresthat the penetration resistance generally increases very slowly within the first two hours aftermixing. At about four to five hours after mixing, however, the penetration resistance startsto increase fairly rapidly indicating that the mortar being tested has then began to solidify andbehave more like a solid.

From the graphs presented in Figs. 4.1 to 4.6, the times taken to reach penetrationresistances of 0.5 MPa and 3.5 MPa are determined and the results are tabulated in Table 4.8.It is evident from these results that the stiffening time of a concrete depends on the grade ofthe concrete, whether PFA has been added and the curing temperature.

- 34 -

Table 4.8 - Stiffening times of the concrete mixes

Batch no.Time to reach penetration

resistance of 0.5 MPa(hour:minute)

Time to reach penetrationresistance of 3.5 MPa

(hour:minute)

G20/0PFA/20C

G20/0PFA/30C

G20/25PFA/20C

G20/25PFA/30C

3:15

3:05

4:45

4:35

4:45

4:10

6:45

6:20

G30/0PFA/20C

G30/0PFA/30C

G30/25PFA/20C

G30/25PFA/30C

3:30

2:40

4:30

3:20

5:10

4:00

6:15

4:50

G40/0PFA/20C

G40/0PFA/30C

G40/25PFA/20C

G40/25PFA/30C

4:15

2:45

4:30

4:15

6:05

4:45

6:30

5:50

- 42 -

expected as the use of PFA to replace part of the cement would generally slow than thestrength development process. Comparing concrete of the same mix composition but curedat different temperatures, it is evident that a higher curing temperature leads to a shorterstiffening time. This can be explained by the fact that cement hydration is generally faster ata higher temperature.

4.3 Material Properties at Time of Hammering Test

For concretes which had hardened sufficiently for the moulds to be removed before thehammering tests, their material properties at the time of conducting hammering test weremeasured. The material properties measured were the cube compressive strength, splitcylinder tensile strength, dynamic modulus, static modulus and longitudinal wave propagationvelocity. Specimens for material properties tests were cast from the same batch of concreteout of which the prisms for hammering test were produced and were cured alongside theprisms so that they were cured under exactly the same conditions. They were tested at thesame time of conducting hammering test and thus the material properties measured may betaken as those of the concrete at the time of hammering. Since the moulds could not beremoved at less than 12 hours after casting, the material properties tests were carried out onlyfor concretes in Series 2 tests, not for concretes in Series 1 tests.

For the determination of cube compressive strength and split cylinder tensile strengthat each designated curing temperature and at each age, three cubes and three cylinders werecast from the concrete. From the three cubes, the cube strength results are averaged to givethe mean compressive strength of the concrete at the time of hammering test. Similarly,from the three cylinders, the split cylinder tensile strength results are averaged to give themean tensile strength of the concrete.

On the other hand, the dynamic modulus and longitudinal wave propagation velocityof the concrete were measured by ultrasonic testing of the concrete prisms cast for thehammering tests. The ultrasonic tests were carried out on the concrete prisms in theirlongitudinal direction just before they were subjected to the hammering impact. Since ineach set of five prisms to be tested for their vibration resistance, only four were subjected tohammer blows, the ultrasonic tests were carried out on the four prisms which would behammered. The four ultrasonic test results from each set of prisms are averaged to give thedynamic modulus and longitudinal wave velocity of the concrete at the time of hammeringtest. Occasionally, especially when the concrete to be tested was at the age of 12 hours, thesignal received from the ultrasonic transducer was too weak for accurate measurement to bemade and consequently no ultrasonic test result was obtained.

The results of the material properties tests are listed in Tables 4.9 to 4.11. For easierinterpretation, the compressive and tensile strengths of the concretes are plotted against age inFigs. 4.8 to 4.13. From these results, it is evident that the curing temperature has asignificant effect on the rate of strength development. Generally speaking, the compressiveand tensile strengths attained by a concrete at a higher curing temperature of 30ºC are higherthan the corresponding strengths attained at a curing temperature of 20ºC at all ages.Moreover, this effect of the curing temperature is somewhat greater for concretes containingPFA.

- 46 -

Table 4.9 - Material properties of G20 concretes at the time of hammering test



Meancompressive

strength(MPa)

Meantensile

strength(MPa)

Dynamicmodulus

(GPa)

Longitudinalwave

velocity(m/s)

12 hrs 3.9 0.43 - -

18 hrs 5.8 0.74 19.4 3140

G20/0PFA/20C 24 hrs 8.9 1.00 22.9 3410

3 days 12.7 1.31 26.9 3700

7 days 17.7 1.54 30.6 3940

28 days 31.9 2.46 34.7 4200

12 hrs 4.8 0.48 15.2 2780

18 hrs 6.3 0.81 19.5 3150

G20/0PFA/30C 24 hrs 11.0 1.21 22.5 3380

3 days 14.7 1.56 23.6 3460

7 days 21.5 1.99 28.1 3780

28 days 32.3 2.68 35.1 4220

12 hrs 1.3 0.13 8.4 2070

18 hrs 3.1 0.28 14.3 2700

G20/25PFA/20C 24 hrs 4.9 0.49 16.4 2890

3 days 8.9 1.01 21.6 3310

7 days 10.5 1.37 25.9 3630

28 days 22.2 1.76 31.8 4020

12 hrs 3.0 0.28 7.8 1990

18 hrs 3.4 0.32 12.9 2560

G20/25PFA/30C 24 hrs 5.0 0.61 16.8 2920

3 days 9.9 1.09 21.6 3310

7 days 12.2 1.08 26.8 3690

28 days 24.5 2.18 31.3 3990

- 47 -




Meancompressive

strength(MPa)

Meantensile

strength(MPa)

Dynamicmodulus

(GPa)

Longitudinalwave

velocity(m/s)

12 hrs 4.1 0.40 13.9 2640

18 hrs 7.5 0.76 18.9 3080

G30/0PFA/20C 24 hrs 10.2 0.94 20.4 3200

3 days 18.6 1.82 28.6 3790

7 days 28.7 2.09 32.1 4010

28 days 39.3 2.87 36.5 4280

12 hrs 5.9 0.52 16.3 2860

18 hrs 9.0 0.81 20.2 3180

G30/0PFA/30C 24 hrs 10.4 0.99 20.9 3240

3 days 21.9 1.79 27.7 3730

7 days 37.1 2.72 33.8 4120

28 days 41.9 3.07 35.9 4240

12 hrs 2.6 0.27 12.2 2490

18 hrs 4.8 0.41 15.9 2840

G30/25PFA/20C 24 hrs 7.4 0.75 19.0 3110

3 days 10.4 1.09 24.7 3540

7 days 18.9 1.62 29.3 3860

28 days 33.9 2.54 34.9 4210

12 hrs 3.6 0.37 12.6 2530

18 hrs 7.2 0.62 18.8 3090

G30/25PFA/30C 24 hrs 8.6 0.77 20.1 3200

3 days 14.8 1.30 27.2 3720

7 days 24.4 2.20 30.1 3910

28 days 43.1 3.04 35.7 4260

- 48 -




Meancompressive

strength(MPa)

Meantensile

strength(MPa)

Dynamicmodulus

(GPa)

Longitudinalwave

velocity(m/s)

12 hrs 2.0 0.29 10.3 2290

18 hrs 5.6 0.66 21.5 3310

G40/0PFA/20C 24 hrs 7.1 0.73 23.1 3430

3 days 23.5 1.79 29.0 3840

7 days 31.2 2.40 32.6 4070

28 days 46.9 3.17 38.2 4410

12 hrs 1.6 0.09 - -

18 hrs 13.0 1.06 22.3 3370

G40/0PFA/30C 24 hrs 13.4 1.12 23.0 3420

3 days 27.0 2.08 28.4 3800

7 days 38.9 2.60 31.3 3990

28 days 50.8 3.26 36.0 4280

12 hrs 0.7 0.03 - -

18 hrs 2.0 0.33 8.8 2110

G40/25PFA/20C 24 hrs 3.9 0.64 16.1 2860

3 days 18.0 1.83 27.1 3710

7 days 26.5 2.31 31.3 3990

28 days 40.5 2.65 32.9 4090

12 hrs 2.2 0.20 8.1 2030

18 hrs 4.3 0.37 18.1 3030

G40/25PFA/30C 24 hrs 8.0 0.65 19.8 3170

3 days 18.3 1.52 25.8 3620

7 days 28.0 1.96 29.6 3880

28 days 48.0 3.03 36.6 4310

- 49 -

Table 4.12 - Static moduli of G20 concretes

Batch no. Age at which hammeringtest carried out

Static modulus(GPa)

12 hrs (not measured)

18 hrs 8.5

G20/0PFA/20C 24 hrs 10.3

3 days 15.4

7 days 18.1

28 days 23.1


18 hrs 8.5

G20/0PFA/30C 24 hrs 10.3

3 days 17.4

7 days 16.2

28 days 23.7



G20/25PFA/20C 24 hrs 11.7

3 days 14.2

7 days 15.9

28 days 20.4


18 hrs 10.3

G20/25PFA/30C 24 hrs 9.4

3 days 13.3

7 days 19.2

28 days 23.8

- 50 -



Static modulus(GPa)


18 hrs 14.7

G30/0PFA/20C 24 hrs 13.4

3 days 19.0

7 days 21.5

28 days 23.6


18 hrs 15.2

G30/0PFA/30C 24 hrs 18.0

3 days 19.4

7 days 24.3

28 days 25.7


18 hrs 13.4

G30/25PFA/20C 24 hrs 15.6

3 days 17.6

7 days 19.0

28 days 22.8


18 hrs 14.4

G30/25PFA/30C 24 hrs 17.7

3 days 19.5

7 days 21.6

28 days 25.3

- 51 -



Static modulus(GPa)


18 hrs 14.5

G40/0PFA/20C 24 hrs 17.6

3 days 17.2

7 days 19.5

28 days 26.5


18 hrs 16.4

G40/0PFA/30C 24 hrs 16.8

3 days 18.1

7 days 19.4

28 days 25.6


18 hrs 6.5

G40/25PFA/20C 24 hrs 10.9

3 days 17.9

7 days 23.9

28 days 24.0


18 hrs 15.1

G40/25PFA/30C 24 hrs 18.4

3 days 18.4

7 days 21.2

28 days 25.0

- 61 -

4.4 Results of Hammering Tests and Integrity Tests

Each set of specimens for the hammering tests consisted of five prisms which werecast from the same batch of concrete and cured under the same designated curing temperature.One of the five prisms was treated as a control specimen and was not subjected to anyhammer blow or vibration. Among the other four prisms, one was subjected to a number ofhammer blows with increasing intensity until at least one crack appeared to determine thevibration intensity required to cause cracking. The remaining three prisms were then eachsubjected to one hammer blow in such a way that at least one prism but not all would becracked. In some cases, however, none of the three prisms was cracked after the hammeringtest. In such case, the three prisms were re-hammered until at least one of them was cracked.In some cases, therefore, among the four prisms subjected to hammering, there could be twoor more prisms subjected to more than one hammer blows.

After the hammering tests, the prisms, including the one which acted as controlspecimen, were continued to be cured at their designated curing temperature until the age of28-day by which time they were tested for their integrity in order to evaluate the effects of theshock vibration applied to them. Any possible self-healing was allowed for by placing theprisms with their longitudinal axis vertical during curing so that any cracks formed by theshock vibration were kept closed.

The results of the hammering and integrity tests are presented in Tables 4.15 to 4.22,Tables 4.23 to 4.30 and Tables 4.31 to 4.38 for the G20, G30 and G40 concretes respectively.

When studying the test results, it should be borne in mind that from each batch ofconcrete, fifteen prisms for hammering tests at three different ages (five prisms for each age)were cast (since there were only fifteen moulds, only fifteen prism could be cast from onebatch of concrete). As one control prism was cast for hammering test at each age, there wereall together three control specimens cast from each batch of concrete. Although each controlspecimen corresponded to hammering test at a certain age, the three control specimens castfrom the same batch of concrete were actually identical as they were cured under the sameconditions and were not subjected to any hammering. However, due to random variation ofthe strength properties, the three control specimens yielded somewhat slightly differentintegrity test results. Since the three control specimens were cast from the same batch ofconcrete, it was found difficult to assign any particular control specimen to hammering test atany particular age. To overcome this difficulty, the integrity test results of the three controlspecimens cast from the same batch of concrete were averaged and the results taken as thosefor the hammering tests at the three different ages. That is why in the integrity test resultstabulated, prisms cast for hammering tests at different ages can have the same controlspecimen results (the averaged results of the three control specimens). In fact, this averagingof the control specimen results has the advantage of reducing the statistical variation of thetest results of the control specimens which would be used as a datum when assessing theeffects of the shock vibration applied to the other prisms.

In Tables 4.15 to 4.38, several peak particle velocity results separated by semi-colonsare sometimes given for the same prism. This is because the prism had been subjected tomore than one hammer blow during the hammering test; in the results presented, each peakparticle velocity given is the intensity of the shock vibration applied at one hammer blow.Occasionally, a hyphen is given as the peak particle velocity of the shock vibration applied to

- 62 -

the prism. This means that no result on vibration intensity was obtained for that particularprism during the hammering test. This was due mainly to the tight schedule of the testingprogramme and the fact that many of the tests had to be carried out at night time (sometimesafter mid-night) when there was no technician to help handling the heavy specimens oroperating the equipment. Accidents and/or mistakes were therefore prone to occur.Fortunately, none of the Research Assistants had any of their toes broken but some mistakeshad been made during the hammering tests. In several cases, it was found after thehammering test was completed that the cable linking the accelerometer to the conditioningamplifier had not been properly connected and consequently, no signal was received by thedata acquisition system during the hammering. In one other case, the data file storing thehammering test results was accidentally deleted before it was backed up. However, theworst thing happened in day time when the accelerometer which cost about HK$10,000 wasbroken during one hammering test (hammering test of G40/0PFA/20C at the age of 28 days).As a result, no peak particle velocity result was obtained for that particular whole set ofprisms (see Table 4.33) and the accelerometer had to be replaced by a new one. In anothercase of G30/25PFA/20C subjected to hammering test at the age of 8 hours, one prism (prismno.4) was damaged accidentally during the direct tensile test and no result for its tensilestrength was thus obtained; this is indicated by a question mark in the corresponding dataentry in Table 4.27. Nevertheless, the number of prisms affected is actually quite small,compared to the total number of prisms tested.

After the application of each hammer blow to the test prism, the prism was inspectedfor any transverse cracks formed on the concrete. Transverse cracks observed are denotedeither as “*” or “#” against the peak particle velocity of the shock wave applied in Tables 4.15- 4.38. A “*” indicates that a transverse crack cutting through the whole section of theconcrete prism had been formed, i.e. the concrete prism had been broken into two pieces. A“#” indicates that one or more transverse cracks on the surface of the concrete prism had beenobserved but the concrete prism still remained as one piece, i.e. the transverse cracks formedhad not cut through the whole section of the prism. Although the exact locations of thetransverse cracks were not measured, they were all found to be located within the middle-thirdof the length of the prisms. This reveals that the tensile stresses induced by the shock waveapplied were highest near the middle of the prisms.

The hammering test results are quite erratic and therefore not easy to interpret. Sinceit was not possible to continuously adjust the vibration intensity until the concrete failed, thevibration resistance of the concrete could not be directly obtained. Nevertheless, from theintensity of the shock vibration applied which had not caused any damage to the concrete andthe intensity of the shock vibration applied which had caused damage to the concrete, thevibration resistance of the concrete may be indirectly determined. Detailed analysis of thetest results is presented in the next two chapters.

- 63 -

Table 4.15 - Hammering and integrity test results for G20/0PFA/20C - Series 1 tests

Age athammer

test

Prismno.

Peak particlevelocity ofshock wave

applied(mm/s)

Directtensile

strength ofhammered

prism at 28-day (MPa)

Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 150 2.13 34.0

2 hrs 2 230 1.44 1.99 36.4 30.9

3 40 2.04 34.3

4 330 1.96 32.9

1 240 1.70 28.8

4 hrs 2 170 2.20 1.99 25.0 30.9

3 380 1.76 32.4

4 95 1.90 32.9

1 110 2.10 33.2

6 hrs 2 290 1.30 1.99 32.2 30.9

3 300 2.02 32.0

4 140 2.16 32.8

1 420 1.82 29.3

8 hrs 2 410 2.11 1.67 32.5 27.0

3 140 2.27 31.6

4 180 2.02 29.8

1 480 1.96 32.1

10 hrs 2 250 2.04 1.67 29.6 27.0

3 160 1.89 28.4

4 290 1.93 29.0

1 150 2.11 29.7

12 hrs 2 230 2.02 1.67 29.7 27.0

3 410 2.11 32.1

4 200 2.09 31.0

- 64 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 21;90 1.76 31.1

2 hrs 2 340 2.22 1.82 30.9 28.8

3 240 1.78 28.0

4 550 1.91 30.2

1 86;120;160 1.49 30.2

4 hrs 2 91 0.56 1.96 29.0 28.3

3 79 1.11 29.1

4 56 1.87 29.9

1 74 1.93 29.9

6 hrs 2 180 1.67 1.96 32.9 28.3

3 110 1.87 30.8

4 79 2.13 31.1

1 170 1.80 28.8

8 hrs 2 140 1.91 1.82 27.9 28.8

3 140 1.93 27.7

4 180 1.78 27.8

1 820 1.69 29.5

10 hrs 2 460;800 2.13 1.99 28.7 28.2

3 750 2.24 29.9

4 680 1.87 29.1

1 740 2.04 28.0

12 hrs 2 650 2.07 1.92 29.2 27.8

3 760;860 1.91 31.0

4 510;640 1.64 29.6

- 65 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 460;600;680# 0.44 29.7

12 hrs 2 640 1.89 2.07 29.6 28.6

3 610* 0.00 28.9

4 570 0.51 30.4

1 670;910;960* 0.00 26.7

18 hrs 2 590 1.67 2.07 26.5 28.6

3 920 1.89 27.8

4 890# 0.18 25.5

1 120;170;420* 0.00 34.4

24 hrs 2 670 2.36 2.00 33.2 33.9

3 350* 0.00 32.9

4 410;700 1.93 33.2

1 1500 1.24 28.7

3 days 2 840;1700# 0.13 2.07 28.3 28.6

3 1500 1.87 29.3

4 700;1100;1200* 0.00 30.2

1 1200 2.44 32.6

7 days 2 1100 2.04 2.00 31.1 33.9

3 1100 2.29 33.2

4 1100;1300;1600# 0.00 33.1

1 1800;1900;3600* 0.00 33.0

28 days 2 1500;1700 2.11 2.00 32.3 33.9

3 1800;2000;2300 2.11 33.8

4 2300# 1.33 32.7

- 66 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 780;1100;1200# 0.00 29.0

12 hrs 2 - 1.60 1.92 27.2 27.8

3 510 1.80 30.1

4 580;600;910 1.80 29.8

1 660;730;990* 0.00 30.1

18 hrs 2 780 1.89 1.82 29.8 28.8

3 810;830;1600* 0.00 31.0

4 1300 2.02 29.7

1 710 2.27 33.4

24 hrs 2 740;990;1300* 0.00 2.19 33.2 29.7

3 840 1.91 32.2

4 700;1100;1200* 0.00 33.3

1 2300;2500;2900* 0.00 29.6

3 days 2 900;1000;2100 2.00 1.99 29.7 28.3

3 1700;2100;2700* 0.00 28.8

4 1100;2100;2500 1.58 28.5

1 600;1800;2900* 0.00 30.7

7 days 2 1200 2.11 2.19 28.7 29.7

3 2300 2.00 29.3

4 3100 2.04 31.1

1 3400 1.89 30.1

28 days 2 2100 1.89 2.19 30.3 29.7

3 1800;2100;2900* 0.00 31.7

4 3300 1.73 31.3

- 67 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 - 1.60 25.3

2 hrs 2 28 1.10 1.55 24.8 20.9

3 67 1.60 25.3

4 35 1.70 25.2

1 150 1.56 26.1

4 hrs 2 70 1.67 1.67 21.7 21.2

3 300 1.47 24.3

4 700 1.60 24.5

1 130 1.13 22.7

6 hrs 2 60 1.62 1.67 22.2 21.2

3 270 1.47 19.7

4 210 0.71 22.4

1 200 1.64 22.2

8 hrs 2 150 1.78 1.60 20.1 24.4

3 1700 0.53 20.7

4 1700 1.38 22.9

1 560 1.73 22.5

10 hrs 2 520 0.27 1.71 21.3 24.6

3 1800 0.98 18.4

4 2100 1.60 20.5

1 660 0.93 20.7

12 hrs 2 560 1.38 1.71 21.2 24.6

3 1600 1.11 22.9

4 3400 1.22 21.6

- 68 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 23 1.96 27.6

2 hrs 2 31 1.69 1.57 27.3 24.4

3 25 1.82 24.2

4 56 1.76 24.8

1 36;38 1.98 31.6

4 hrs 2 36 1.91 1.73 34.2 29.0

3 150 1.87 33.3

4 140 1.76 34.2

1 210;310 1.89 32.2

6 hrs 2 180;210 2.00 1.73 33.7 29.0

3 170 1.93 32.8

4 250 1.87 32.7

1 230 1.44 26.3

8 hrs 2 140 1.84 1.57 25.9 24.4

3 580 1.89 26.9

4 1100 1.69 26.9

1 290;300;380 0.93 29.6

10 hrs 2 360;410 1.67 1.91 29.9 28.1

3 370 1.42 28.9

4 310 0.51 28.3

1 300;310 1.67 27.8

12 hrs 2 310;370;400 1.13 1.91 29.6 28.1

3 250 1.76 30.2

4 360;460 0.22 30.8

- 69 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 340# 0.00 20.6

12 hrs 2 980 1.09 1.71 20.6 24.6

3 -* 0.00 20.6

4 -* 0.00 21.4

1 510;1200# 0.98 20.2

18 hrs 2 520;920 1.24 1.67 20.1 21.2

3 2000;2300 1.18 20.6

4 1600 1.38 21.6

1 550;750# 0.90 22.4

24 hrs 2 510;620 1.60 1.55 22.0 20.9

3 480;1200 0.90 21.0

4 760;1200;1500# 0.60 21.9

1 1400 1.47 18.7

3 days 2 1100 1.33 1.60 21.9 24.4

3 2800* 0.00 22.6

4 2300# 0.76 21.3

1 2800* 0.00 23.3

7 days 2 2200* 0.00 1.70 22.9 21.7

3 1000 0.93 22.9

4 2600 1.70 21.7

1 3100 1.60 23.7

28 days 2 1400 1.60 1.70 24.1 21.7

3 2300# 1.10 23.0

4 2200 1.60 23.4

- 70 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 450;570;700 1.42 30.3

12 hrs 2 440 1.67 1.91 28.2 28.1

3 390 1.76 29.8

4 450 1.56 29.5

1 700 1.62 29.8

18 hrs 2 590;650;820 1.29 1.73 28.2 29.0

3 550;880 1.18 29.3

4 930 1.87 28.6

1 640;710;1500 1.40 23.1

24 hrs 2 860 1.20 1.66 23.7 24.3

3 740 1.36 23.6

4 1100;1400;1700 1.60 24.2

1 2600# 0.71 24.9

3 days 2 1200 0.62 1.57 24.0 24.4

3 2200* 0.00 22.5

4 2500* 0.00 24.8

1 3500# 0.33 26.5

7 days 2 2300 1.27 1.66 26.3 24.3

3 1800 1.91 24.4

4 1700 1.62 26.4

1 3100* 0.00 26.1

28 days 2 2900 1.02 1.66 25.7 24.3

3 3200# 0.56 25.6

4 2900* 0.00 26.0

- 71 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 24;100 2.31 40.8

2 hrs 2 100 2.18 2.13 41.9 39.7

3 92 2.16 42.1

4 - 2.02 40.4

1 74;88 2.07 39.8

4 hrs 2 320 1.82 2.13 35.4 39.7

3 410 2.33 34.0

4 500 2.00 37.5

1 100;180 1.69 32.2

6 hrs 2 160;240 1.98 1.89 35.8 39.3

3 600 1.89 36.3

4 670 1.80 33.5

1 180;280 2.11 34.6

8 hrs 2 340;390 1.80 1.93 38.7 39.6

3 500 1.13 38.4

4 490 1.89 37.5

1 220;330;400 1.89 39.4

10 hrs 2 410 1.22 1.89 40.8 41.2

3 370;540 2.24 41.0

4 350 2.16 40.2

1 490;560 1.44 41.5

12 hrs 2 500;820 2.20 1.93 38.5 41.5

3 300 2.02 40.0

4 530 1.67 41.0

- 72 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 170;180 2.07 44.0

2 hrs 2 150 2.18 2.09 43.3 42.3

3 46 2.02 45.0

4 29 1.78 41.5

1 120;260;280 2.04 44.6

4 hrs 2 170 1.80 2.09 38.6 42.3

3 230 1.89 40.4

4 300 1.96 41.1

1 340;350;440 1.98 37.5

6 hrs 2 210;230 2.18 2.11 42.7 40.8

3 210 2.07 39.1

4 310 2.20 40.9

1 340;410;430 2.11 41.3

8 hrs 2 600 1.84 2.11 40.5 40.8

3 350 1.89 40.6

4 390 2.16 38.5

1 570;700;770 1.24 37.7

10 hrs 2 560 1.89 1.98 38.2 42.8

3 610;830 1.69 39.4

4 710 2.36 39.9

1 1100;1200 2.20 37.8

12 hrs 2 430;780 2.00 1.98 39.9 42.8

3 780 1.67 39.0

4 980 2.07 38.2

- 73 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 310;460;800# 0.71 41.2

12 hrs 2 760 2.00 1.93 40.8 41.2

3 510 1.87 40.0

4 730 1.80 40.4

1 810;1000* 0.00 37.5

18 hrs 2 430 1.89 2.13 38.7 39.7

3 700 2.02 38.6

4 730;1300# 0.38 37.2

1 980;1400;1600# 0.53 38.7

24 hrs 2 1600 1.36 1.93 37.5 40.0

3 1000 1.71 38.2

4 990;1100# 0.36 37.2

1 890;1100;1100* 0.00 36.4

3 days 2 1000# 0.36 2.02 38.1 39.9

3 620 1.89 40.4

4 740 2.58 38.7

1 1300;1400;1600* 0.00 37.4

7 days 2 1700# 0.47 2.02 36.6 39.9

3 1100 1.89 38.3

4 950 2.33 39.9

1 1500;1800;1900* 0.00 38.6

28 days 2 1400* 0.00 2.02 38.6 39.9

3 1200 1.80 38.5

4 1100;1500 0.47 38.1

- 74 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 590;800;1100# 0.36 37.8

12 hrs 2 1600 2.09 1.98 41.0 42.8

3 810 2.00 39.8

4 1900* 0.00 36.5

1 990 1.67 39.2

18 hrs 2 -* 0.00 2.09 38.6 42.3

3 900;1200# 0.38 36.8

4 690 1.80 40.3

1 1500;1600* 0.00 40.5

24 hrs 2 1500 1.87 2.11 39.5 40.8

3 940 1.56 39.7

4 1400;1500# 0.22 39.4

1 1300;1400* 0.00 40.0

3 days 2 1000;1500* 0.00 2.04 39.5 41.0

3 980 2.11 40.5

4 1100 1.93 37.7

1 1500;1700# 0.42 38.1

7 days 2 1500;1600* 0.00 2.04 35.7 41.0

3 720 1.71 39.6

4 1700 1.20 38.0

1 1100;1500;1800* 0.00 38.6

28 days 2 1400# 0.38 2.04 39.2 41.0

3 1500# 0.67 39.7

4 1900* 0.00 39.8

- 75 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 22;90 1.60 28.8

2 hrs 2 22 1.33 1.49 27.6 29.5

3 30 1.51 29.7

4 32 1.13 26.7

1 230;440 1.62 31.4

4 hrs 2 490 1.69 1.71 31.9 32.7

3 280 1.80 33.5

4 620 1.62 32.0

1 200 1.64 32.8

6 hrs 2 84;190;260 2.16 1.71 33.6 32.7

3 290 1.78 31.6

4 410 1.89 32.8

1 210;250;280 1.89 34.0

8 hrs 2 380 1.60 1.71 33.6 32.7

3 320 1.78 31.7

4 400;480 ? 33.9

1 340;400;540 1.42 28.7

10 hrs 2 620 1.33 1.49 27.9 29.5

3 460 1.58 30.5

4 650;790 1.31 29.2

1 260;480;560 1.42 29.1

12 hrs 2 370 1.56 1.49 29.3 29.5

3 470 1.29 27.4

4 860 1.44 28.9

- 76 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 20;47 1.76 40.8

2 hrs 2 36 1.93 1.71 43.1 38.8

3 29 2.02 40.1

4 69 1.80 41.4

1 180;240;350 1.93 40.7

4 hrs 2 460 1.84 1.71 39.4 39.0

3 240 1.73 38.4

4 370 1.84 39.6

1 410;490;510 1.42 37.8

6 hrs 2 670 1.40 1.71 38.7 39.0

3 630 1.80 39.4

4 660 1.11 37.7

1 520;960;1100 2.11 40.4

8 hrs 2 860 1.82 1.71 37.5 39.0

3 790 1.64 38.4

4 860 0.71 39.2

1 440;640;660 1.89 40.0

10 hrs 2 760 1.71 1.71 39.2 38.8

3 450 1.60 38.6

4 730 1.78 38.6

1 550;560;830 1.76 34.2

12 hrs 2 580 1.82 1.71 38.7 38.8

3 500 1.78 37.3

4 670 1.96 35.5

- 77 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 550;610;1100# 0.22 31.2

12 hrs 2 680 1.71 1.87 31.5 33.1

3 580 1.78 33.8

4 1000 1.56 31.8

1 750;850# 0.40 37.7

18 hrs 2 760 1.76 1.62 36.6 34.8

3 1000 1.02 34.1

4 1000 1.82 35.2

1 870;1300# 0.47 32.9

24 hrs 2 1300 1.62 1.87 31.0 33.1

3 890 1.69 31.8

4 1500* 0.00 31.2

1 1100;1300;2000* 0.00 32.4

3 days 2 1400 1.76 1.62 30.7 34.8

3 940 1.44 34.5

4 1700* 0.00 34.0

1 1300;1400* 0.00 32.2

7 days 2 1200# 0.40 1.87 31.1 33.1

3 810 1.71 32.7

4 1400;1500;1700* 0.00 32.8

1 1600;1900;1900 1.27 34.0

28 days 2 1400;1600;1700# 0.00 1.62 33.7 34.8

3 1400 1.44 33.4

4 1500# 0.56 33.4

- 78 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 680;700 1.38 35.5

12 hrs 2 430;640* 0.00 1.78 33.8 38.5

3 450 1.67 36.8

4 410 1.44 35.1

1 910;1100;1400# 0.38 38.1

18 hrs 2 980 1.04 1.78 40.5 38.5

3 780 1.71 39.1

4 1300# 0.33 39.2

1 1300;1600;1700# 0.71 39.5

24 hrs 2 1200 1.09 1.80 38.4 39.1

3 1300 1.82 40.2

4 1700# 0.36 38.4

1 1300;1400# 0.33 33.3

3 days 2 1700* 0.00 1.78 31.6 38.5

3 900 1.44 38.7

4 1500;1600# 0.33 42.9

1 1300;1400;2000# 0.56 38.7

7 days 2 1300 0.40 1.80 35.9 39.1

3 1300 1.64 37.4

4 1500;1600* 0.00 38.2

1 1700;2000;2200* 0.00 39.7

28 days 2 1600 0.78 1.80 39.5 39.1

3 1400* 0.00 39.1

4 980 1.71 39.1

- 79 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 49;98;110 2.36 46.8

2 hrs 2 38;45;53 2.53 2.84 49.1 47.0

3 28;88 2.18 50.0

4 100;130 2.36 51.2

1 74;80 2.49 52.5

4 hrs 2 50;110 2.13 2.40 51.6 50.0

3 74;98;270 2.36 54.9

4 150;180;190 1.47 51.9

1 290;350;420 2.58 51.1

6 hrs 2 - 1.78 2.11 49.2 46.6

3 100;110;130 1.89 44.1

4 110 2.33 47.1

1 130;150;330 2.44 52.1

8 hrs 2 150;210;220 2.56 2.47 50.1 49.9

3 160;180;180 2.42 53.3

4 430;450 2.44 50.3

1 230;290;300 2.42 49.5

10 hrs 2 210;250 2.51 2.11 54.5 46.6

3 170;380 2.53 58.3

4 240;390 2.38 58.4

1 210;210;240 2.44 52.4

12 hrs 2 230;390 2.27 2.40 51.9 50.0

3 200;200;210 2.67 51.2

4 320;420 1.96 51.4

- 80 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 55;68;97 1.87 46.2

2 hrs 2 91 2.58 2.36 47.7 48.3

3 27;42 2.13 47.6

4 61 2.16 43.0

1 74;100 2.02 46.7

4 hrs 2 45;95 2.31 2.31 53.4 47.1

3 110 2.16 52.6

4 94;110 2.24 55.4

1 82;160;190 2.56 41.6

6 hrs 2 45 2.13 2.44 45.3 50.3

3 96;150 1.98 44.3

4 87;140 2.67 43.0

1 100;250 1.58 47.4

8 hrs 2 180 1.89 2.22 50.7 46.9

3 42;110 1.60 44.1

4 41;200 2.22 52.3

1 180;260 2.42 45.3

10 hrs 2 160 2.44 2.44 46.0 50.3

3 190 1.13 50.1

4 200 2.00 44.4

1 170;290;320 1.67 52.9

12 hrs 2 370 1.82 2.31 48.3 47.1

3 250;390 1.89 47.2

4 240 2.24 51.2

- 81 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 380;430;520 1.49 46.8

12 hrs 2 1500* 0.00 2.47 45.7 49.9

3 680;880# 0.87 49.0

4 580;970;1100 1.00 49.9

1 840;930;1300 2.31 47.9

18 hrs 2 420;560;890 2.58 2.84 47.4 47.0

3 870;1000* 0.00 48.6

4 320 2.22 47.5

1 1400;1900 1.22 50.9

24 hrs 2 1300 1.40 2.47 52.1 49.9

3 2100* 0.00 50.1

4 720 2.00 50.5

1 270;590;610* 0.00 48.6

3 days 2 230;400;500 2.44 2.84 47.5 47.0

3 350;710* 0.00 47.9

4 210;300 2.31 48.8

1 240 1.89 45.9

7 days 2 940# 1.09 2.11 49.5 46.6

3 500;1100;1400* 0.00 41.3

4 910 2.13 47.2

1 -* 0.00 50.2

28 days 2 - 2.04 2.40 50.1 50.0

3 -* 0.00 47.7

4 - 1.38 49.7

- 82 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 370;380;500* 0.00 45.2

12 hrs 2 380* 0.00 2.22 45.1 46.9

3 510 1.60 44.1

4 280 1.91 47.3

1 490;550* 0.00 46.3

18 hrs 2 330;560* 0.00 2.36 47.0 48.3

3 330 1.60 43.9

4 180 1.91 46.0

1 750;920# 0.93 43.8

24 hrs 2 1300* 0.00 2.22 50.2 46.9

3 1200 1.60 46.2

4 1100 2.22 47.5

1 - 2.18 50.0

3 days 2 250;850;1800* 0.00 2.36 43.0 48.3

3 - 2.16 47.6

4 -* 0.00 46.3

1 800;990;1000* 0.00 47.4

7 days 2 1100;1600# 0.00 2.44 43.5 50.3

3 690 2.33 48.2

4 2200* 0.00 47.8

1 570;830 2.02 46.7

28 days 2 790* 0.00 2.31 53.4 47.1

3 -* 0.00 52.6

4 - 2.24 55.3

- 83 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 76;78 1.98 42.2

2 hrs 2 110;200;320 2.36 2.00 42.2 41.5

3 74 1.36 42.6

4 79 2.00 42.3

1 63;70 2.60 42.8

4 hrs 2 73;120 0.36 2.00 44.1 41.5

3 52 0.91 43.3

4 50 1.78 43.3

1 60;67 1.91 46.6

6 hrs 2 170;200 2.40 2.04 46.4 43.9

3 96 1.60 48.0

4 120 1.96 47.0

1 81;81 2.27 47.3

8 hrs 2 62;79 1.56 2.11 49.2 45.3

3 150 1.82 44.2

4 180 1.87 46.9

1 81;120 2.27 41.6

10 hrs 2 150;170 2.53 2.49 40.4 40.6

3 110 2.51 40.2

4 120 2.42 40.7

1 170;170 2.44 39.1

12 hrs 2 160;160 2.22 2.49 39.7 40.6

3 260 2.20 41.1

4 350# 0.00 44.5

- 84 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 50;88 2.53 46.3

2 hrs 2 46 1.44 2.27 48.8 47.4

3 30 1.91 47.5

4 35 1.73 47.6

1 170;250 2.36 52.3

4 hrs 2 69 2.60 2.27 51.7 47.4

3 95 1.62 44.3

4 210 2.24 49.4

1 74;150 2.78 46.8

6 hrs 2 130;140 2.44 2.26 45.8 48.7

3 200 2.16 46.8

4 200 2.47 48.1

1 210 2.29 50.7

8 hrs 2 170 1.62 2.26 48.7 48.7

3 220 2.08 49.5

4 230 2.24 49.1

1 390;420 2.49 52.3

10 hrs 2 530 1.38 2.47 47.9 49.0

3 530 2.07 50.8

4 340 2.71 49.9

1 410;550 2.44 49.1

12 hrs 2 550 2.13 2.47 51.6 49.0

3 700 2.18 48.8

4 240 1.84 50.3

- 85 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 110;300;520 1.98 41.7

12 hrs 2 180;180;400 1.69 2.49 39.8 40.6

3 73 2.22 41.3

4 250 1.69 39.9

1 220;250;320* 0.00 44.1

18 hrs 2 480 1.80 2.11 42.9 45.3

3 310# 0.53 43.1

4 510 1.87 43.3

1 410 1.98 44.0

24 hrs 2 230;360* 0.00 2.00 41.9 41.5

3 550;570# 0.91 40.4

4 560;620 2.44 38.1

1 1300 1.47 42.1

3 days 2 690;750;920 1.58 2.04 43.5 43.9

3 750;840;1100# 0.36 40.8

4 860 1.56 43.7

1 830;1000;1300* 0.00 43.6

7 days 2 810;1200 0.58 2.11 42.8 45.3

3 1200 0.67 44.2

4 1800* 0.00 43.6

1 2100* 0.00 39.5

28 days 2 1500 0.67 2.04 39.0 43.9

3 1400* 0.00 40.5

4 750 1.80 39.6

- 86 -


Age athammer

test

Prismno.


applied(mm/s)

Directtensile

strength ofhammered


Directtensile

strength ofcontrol


Equivalentcube

strength ofhammered


Equivalentcube

strength ofcontrol


1 590;600 1.84 50.0

12 hrs 2 560 2.22 2.47 46.1 49.0

3 480# 0.98 47.7

4 580 1.87 47.9

1 810;960;1000# 0.71 50.8

18 hrs 2 850;970;1100 1.09 2.20 47.5 45.2

3 1000 2.36 47.4

4 1000 2.36 51.1

1 1100;1100# 0.38 48.4

24 hrs 2 860# 0.67 2.26 42.6 48.7

3 990 2.11 45.7

4 1000# 1.29 43.5

1 1100;1800* 0.00 46.1

3 days 2 1700# 0.60 2.27 49.3 47.4

3 1400 1.13 47.6

4 -* 0.00 47.2

1 1600;1700# 0.49 47.3

7 days 2 1600 1.22 2.20 45.9 45.2

3 1500* 0.00 47.3

4 1400# 0.09 48.8

1 1400;2000;2000* 0.00 48.8

28 days 2 2000 1.96 2.20 48.2 45.2

3 1400 2.27 48.1

4 770;1600* 0.00 47.5

- 87 -

5. VIBRATION RESISTANCE OF THE CONCRETE MIXES TESTED

5.1 Effects of Shock Vibration on Concrete

The effects of the shock vibrations applied to the concrete prisms during thehammering test are evaluated by comparing the tensile and compressive strengths of thehammered prisms to those of the control prisms which had not been subjected to anyhammering. For such comparison purpose, the following tensile and compressive strengthratios are defined:

tensile strength ratio:

RP

Pt

t1 =

′…(5.1)

compressive strength ratio:

RP

Pc

c2 =

′…(5.2)

in which ′Pt and ′Pc are the tensile and compressive strengths of the hammered prism at 28-day, and Pt and Pc are the tensile and compressive strengths of the corresponding controlprism at 28-day. A strength ratio smaller/greater than 1.0 indicates that the shock vibrationhad reduced/increased the strength of the concrete. On the other hand, a strength ratio closeto 1.0 indicates that the shock vibration had no significant effect on the strength of theconcrete.

The tensile and compressive strength ratios of the hammered prisms are tabulated inTables 5.1 - 5.24 alongside the peak particle velocities of the shock vibrations applied to theprisms during the hammering tests. From the values of R1 and R2 presented, it can beclearly seen whether the vibrations applied had caused cracking and/or significant change inthe strength of the concrete.

The results in Tables 5.1 - 5.24 reveal that whilst the high intensity shock vibrationsinduced onto the concrete prisms during the hammering tests had caused cracking orsubstantial reduction of the tensile strength of the concrete, they had in general little effect onthe compressive strength of the concrete. Among the 576 concrete prisms which had beensubjected to hammering, only two (Prism No.3 of G20/25PFA/20C tested at 10 hrs. and PrismNo.1 of G20/25PFA/20C tested at 3 days) had their compressive strengths lower than those oftheir respective control prisms by more than 20%. Apart from these two prisms, all the otherhammered prisms, representing more than 99% of the total number, had their compressivestrengths remaining at above 80% of those of the respective control prisms. In fact, some ofthe hammered prisms even have higher compressive strength than the respective controlprisms, i.e. values of R2 higher than 1.0. It is not believed that the shock vibrations couldincrease the compressive strength of the concrete, especially after the concrete had hardened.The somewhat higher compressive strength of the hammered prisms than the control prismswas most likely just the result of experimental errors and random variation of concrete

- 88 -

strength. For the same reason, it is suspected that the low R2 values of the two prisms withR2 smaller than 0.8 were partly due to experimental errors and random variation of concretestrength, rather than entirely the effect of the shock vibrations applied. Paying particularattention to those prisms which had been broken into pieces by the shock vibrations applied, itcan be seen that even when vibrations strong enough to break the concrete into pieces hadbeen applied, the compressive strength of the broken pieces of the prisms did not seem tohave been significantly affected.

It is obvious from the test results that the major effect of the shock vibrations appliedto the concrete prisms was on the tensile strength of the concrete. Depending on theintensity of the vibration applied, the shock vibrations had caused the following differentdegrees of damage to the prisms:

(1) the concrete prism was completely cracked and thus broken into two pieces, i.e. theconcrete prism had cracks formed cutting through a whole section of it;

(2) the concrete prism was partially cracked in such a way that hairline cracks observableon the surface had been formed but the concrete prism had remained in one piece, i.e.the cracks formed had not cut through any whole section of the prism;

(3) the concrete prism had remained un-cracked (to be more specific, the concrete prismdid not have any observable cracks formed on its surfaces) but its tensile strength hadbeen significant reduced, as reflected by a relatively low R1 value;

(4) the concrete prism had not been affected in any way, i.e. there was no observablecrack formed and the concrete prism behaved just like the control prism when itstensile and compressive strengths were measured.

In Tables 5.1 - 5.24, prisms which had been completely or partially cracked are marked with“*” or “#” respectively against the shock vibration which caused the cracking.

From the results on cracking, it is seen that the effect of the shock vibrations appliedwas quite erratic. In many cases, two prisms cast from the same concrete mix and cured sideby side under exactly the same conditions were found to be affected to very different extentwhen subjected to shock vibrations of similar intensity (e.g. Prisms No.2 and No.3 ofG20/0PFA/20C tested at 6 hrs.). Quite often, one prism was broken into pieces whensubjected to shock vibration of a certain intensity while another prism, which should beidentical, withstood a much higher intensity vibration without cracking or any deterioration instrength (e.g. Prisms No.3 and No.4 of G20/0PFA/20C tested at 24 hrs.).

In order to allow for any possible self-healing effects, the hammered prisms werecontinued to be cured with their longitudinal axis oriented vertically so that any transversecracks formed were kept closed under the self weight of the prisms. Prisms, which werecompletely cracked and broken into pieces, were reassembled by putting the broken piecesback together before curing. It is, however, found from the results presented in Tables 5.1 -5.24 that all the prisms which had been completely cracked (those with a “*” marked) hadnegligible tensile strength when tested at 28-day, despite continuous curing with the cracksclosed. This indicates that once the concrete has been completely cracked, there is littlehope that any significant tensile strength can be recovered, i.e. there is practically no self-healing that can be relied on to reinstate the cracks. Regarding the prisms which had onlybeen partially cracked (those with a “#” marked), their tensile strengths at 28-day ranged from0.00 to 0.67 of those of the respectively control prisms. It is hard to say whether significant

- 89 -

self-healing had occurred because the tensile strengths of the partially cracked prisms at 28-day might just be constituted largely of the residual strength of the cracked concrete. In anycase, even with the cracks kept closed and the concrete continued to be cured so that anypossible self-healing should have taken place, the tensile strengths of the partially crackedconcrete prisms were found to be all lower than 70% of those of the respective control prisms.Thus, it may be said that once the concrete has observable cracks formed, it is unlikely itstensile strength can be recovered to more than 70% of that of the original concrete.

In each set of four prisms cast from the same concrete and subjected to hammering testat the same age, some prisms were each subjected to only one hammer blow while the otherprisms were each subjected to two or three hammer blows. However, since no directcomparison can be made between prisms subjected to only one hammer blow and prismssubjected to two or three hammer blows (the intensities of the hammer blows applied to themwere quite different), it is difficult to single out the cumulative effect of multiple hammerblows. Added with the uncertainties due to the erratic nature of the effects of shockvibrations on concrete, no definite conclusion can be drawn on the cumulative effect from thepresent set of data. To be on the conservative side, the prisms which had been subjected tomore than one hammer blow are treated as if they had been each subjected to just one hammerblow - the hammer blow of the highest intensity among the hammer blows that the prism hadbeen subjected to. As without the cumulative effect, the values of R1 and R2 should havebeen higher, treating the prisms as if they were each subjected to just one hammer blow wouldtend to underestimate the values of R1 and R2 for the effect of one single hammer blow.

Since it is not possible to continuously adjust the shock vibration intensity until theconcrete fails, the vibration resistance of concrete cannot be directly measured. Nevertheless,for this particular study in which shock vibrations of different intensity were applied in such away that some of the test prisms were cracked while the others remained un-cracked, thevibration resistances of the concretes tested may be indirectly determined from the followingtwo peak particle velocity (ppv) results: (1) the lowest ppv applied that had caused failure ofthe concrete; and (2) the highest ppv applied that had not caused any failure of the concrete.Since shock vibrations have little effect on the compressive strength, failure in this casemeans either cracking or significant reduction (say, more than 30% reduction) in tensilestrength. For easy reference, values of R1 lower than 0.70 which imply more than 30%reduction in tensile strength are underlined in Tables 5.1 - 5.24. A reduction in tensilestrength of less than 30% is not regarded as significant in the present context because of thefollowing reasons:

(1) the tensile strength of concrete is by itself a variable property and when its intrinsicvariation is added with the experimental errors involved in its measurement, the measuredtensile strength can vary by as much as 20% or probably more in some extreme cases; and

(2) appropriate safety margins should have been included in any structural design to allow forpossible variations (either expected or unexpected) in the strength of the materials usedand a known reduction in the tensile strength of concrete of less than 30% should bewithin the normal safety margin provided.

The above two indirect measurements of the vibration resistances of the concretemixes tested are abbreviated as ppv1 and ppv2, which are defined below:

- 90 -

ppv1 = lowest ppv applied that had caused failure …(5.3)

ppv2 = highest ppv applied that had not caused failure …(5.4)

These two ppv values are tabulated in the last two columns of Tables 5.1 - 5.24. Sincetheoretically ppv1 should be higher than ppv2, in cases where none of the hammered prismsin the same set had failed, ppv1 is taken to have a value of “>ppv2”, while in cases where allof the hammered prisms in the same set had failed, ppv2 is taken to be equal to “<ppv1”.However, from those sets of prisms in which at least one but not all of the hammered prismsin the same set had failed, it can be seen that due to the erratic nature of the vibrationresistance of concrete, ppv1 is not always greater than ppv2.

- 91 -

Table 5.1 - Effects of shock vibrations on G20/0PFA/20C - Series 1 tests

Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 150 1.07 1.10

2 hrs 2 230 0.72 1.18 >330 330

3 40 1.03 1.11

4 330 0.98 1.06

1 240 0.85 0.93

4 hrs 2 170 1.11 0.81 >380 380

3 380 0.88 1.05

4 95 0.95 1.06

1 110 1.06 1.07

6 hrs 2 290 0.65 1.04 290 300

3 300 1.02 1.04

4 140 1.09 1.06

1 420 1.09 1.09

8 hrs 2 410 1.26 1.20 >420 420

3 140 1.36 1.17

4 180 1.21 1.10

1 480 1.17 1.19

10 hrs 2 250 1.22 1.10 >480 480

3 160 1.13 1.05

4 290 1.16 1.07

1 150 1.26 1.10

12 hrs 2 230 1.21 1.10 >410 410

3 410 1.26 1.19

4 200 1.25 1.15

- 92 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 21;90 0.97 1.08

2 hrs 2 340 1.22 1.07 >550 550

3 240 0.98 0.97

4 550 1.05 1.05

1 86;120;160 0.76 1.07

4 hrs 2 91 0.29 1.02 79 160

3 79 0.57 1.03

4 56 0.95 1.06

1 74 0.98 1.06

6 hrs 2 180 0.85 1.16 >180 180

3 110 0.95 1.09

4 79 1.09 1.10

1 170 0.99 1.00

8 hrs 2 140 1.05 0.97 >180 180

3 140 1.06 0.96

4 180 0.98 0.97

1 820 0.85 1.05

10 hrs 2 460;800 1.07 1.02 >820 820

3 750 1.13 1.06

4 680 0.94 1.03

1 740 1.06 1.01

12 hrs 2 650 1.08 1.05 >860 860

3 760;860 0.99 1.12

4 510;640 0.85 1.06

- 93 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 460;600;680# 0.21 1.04

12 hrs 2 640 0.91 1.03 570 640

3 610* 0.00 1.01

4 570 0.25 1.06

1 670;910;960* 0.00 0.93

18 hrs 2 590 0.81 0.93 890 920

3 920 0.91 0.97

4 890# 0.09 0.89

1 120;170;420* 0.00 1.01

24 hrs 2 670 1.18 0.98 350 700

3 350* 0.00 0.97

4 410;700 0.97 0.98

1 1500 0.60 1.00

3 days 2 840;1700# 0.06 0.99 1200 1500

3 1500 0.90 1.02

4 700;1100;1200* 0.00 1.06

1 1200 1.22 0.96

7 days 2 1100 1.02 0.92 1600 1200

3 1100 1.15 0.98

4 1100;1300;1600# 0.00 0.98

1 1800;1900;3600* 0.00 0.97

28 days 2 1500;1700 1.06 0.95 2300 2300

3 1800;2000;2300 1.06 1.00

4 2300# 0.67 0.96

- 94 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 780;1100;1200# 0.00 1.04

12 hrs 2 - 0.83 0.98 1200 910

3 510 0.94 1.08

4 580;600;910 0.94 1.07

1 660;730;990* 0.00 1.05

18 hrs 2 780 1.04 1.03 990 1300

3 810;830;1600* 0.00 1.08

4 1300 1.11 1.03

1 710 1.04 1.12

24 hrs 2 740;990;1300* 0.00 1.12 1200 840

3 840 0.87 1.08

4 700;1100;1200* 0.00 1.12

1 2300;2500;2900* 0.00 1.05

3 days 2 900;1000;2100 1.01 1.05 2700 2500

3 1700;2100;2700* 0.00 1.02

4 1100;2100;2500 0.79 1.01

1 600;1800;2900* 0.00 1.03

7 days 2 1200 0.96 0.97 2900 3100

3 2300 0.91 0.99

4 3100 0.93 1.05

1 3400 0.86 1.01

28 days 2 2100 0.86 1.02 2900 3400

3 1800;2100;2900* 0.00 1.07

4 3300 0.79 1.05

- 95 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 - 1.03 1.21

2 hrs 2 28 0.71 1.19 >67 67

3 67 1.03 1.21

4 35 1.10 1.21

1 150 0.93 1.23

4 hrs 2 70 1.00 1.02 >700 700

3 300 0.88 1.15

4 700 0.96 1.16

1 130 0.68 1.07

6 hrs 2 60 0.97 1.05 130 270

3 270 0.88 0.93

4 210 0.43 1.06

1 200 1.03 0.91

8 hrs 2 150 1.11 0.82 1700 1700

3 1700 0.33 0.85

4 1700 0.86 0.94

1 560 1.01 0.91

10 hrs 2 520 0.16 0.87 520 2100

3 1800 0.57 0.75

4 2100 0.94 0.83

1 660 0.54 0.84

12 hrs 2 560 0.81 0.86 660 3400

3 1600 0.65 0.93

4 3400 0.71 0.88

- 96 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 23 1.25 1.13

2 hrs 2 31 1.08 1.12 >56 56

3 25 1.16 0.99

4 56 1.12 1.02

1 36;38 1.14 1.09

4 hrs 2 36 1.10 1.18 >150 150

3 150 1.08 1.15

4 140 1.02 1.18

1 210;310 1.09 1.11

6 hrs 2 180;210 1.16 1.16 >310 310

3 170 1.12 1.13

4 250 1.08 1.13

1 230 0.92 1.08

8 hrs 2 140 1.17 1.06 >1100 1100

3 580 1.20 1.10

4 1100 1.08 1.10

1 290;300;380 0.49 1.05

10 hrs 2 360;410 0.87 1.06 310 410

3 370 0.74 1.03

4 310 0.27 1.01

1 300;310 0.87 0.99

12 hrs 2 310;370;400 0.59 1.05 400 310

3 250 0.92 1.07

4 360;460 0.12 1.10

- 97 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 340# 0.00 0.84

12 hrs 2 980 0.64 0.84 340 <340

3 -* 0.00 0.84

4 -* 0.00 0.87

1 510;1200# 0.59 0.95

18 hrs 2 520;920 0.74 0.95 1200 2300

3 2000;2300 0.71 0.97

4 1600 0.83 1.02

1 550;750# 0.58 1.07

24 hrs 2 510;620 1.03 1.05 750 620

3 480;1200 0.58 1.00

4 760;1200;1500# 0.39 1.05

1 1400 0.92 0.77

3 days 2 1100 0.83 0.90 2300 1400

3 2800* 0.00 0.93

4 2300# 0.48 0.87

1 2800* 0.00 1.07

7 days 2 2200* 0.00 1.06 1000 2600

3 1000 0.55 1.06

4 2600 1.00 1.00

1 3100 0.94 1.09

28 days 2 1400 0.94 1.11 2300 3100

3 2300# 0.65 1.06

4 2200 0.94 1.08

- 98 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 450;570;700 0.74 1.08

12 hrs 2 440 0.87 1.00 >700 700

3 390 0.92 1.06

4 450 0.82 1.05

1 700 0.94 1.03

18 hrs 2 590;650;820 0.75 0.97 880 930

3 550;880 0.68 1.01

4 930 1.08 0.99

1 640;710;1500 0.84 0.95

24 hrs 2 860 0.72 0.98 >1700 1700

3 740 0.82 0.97

4 1100;1400;1700 0.96 1.00

1 2600# 0.45 1.02

3 days 2 1200 0.39 0.98 1200 <1200

3 2200* 0.00 0.92

4 2500* 0.00 1.02

1 3500# 0.20 1.09

7 days 2 2300 0.77 1.08 3500 2300

3 1800 1.15 1.00

4 1700 0.98 1.09

1 3100* 0.00 1.07

28 days 2 2900 0.61 1.06 2900 <2900

3 3200# 0.34 1.05

4 2900* 0.00 1.07

- 99 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 24;100 1.08 1.03

2 hrs 2 100 1.02 1.06 >100 100

3 92 1.01 1.06

4 - 0.95 1.02

1 74;88 0.97 1.00

4 hrs 2 320 0.85 0.89 >500 500

3 410 1.09 0.86

4 500 0.94 0.94

1 100;180 0.89 0.82

6 hrs 2 160;240 1.05 0.91 >670 670

3 600 1.00 0.92

4 670 0.95 0.85

1 180;280 1.09 0.87

8 hrs 2 340;390 0.93 0.98 500 490

3 500 0.59 0.97

4 490 0.98 0.95

1 220;330;400 1.00 0.96

10 hrs 2 410 0.65 0.99 410 540

3 370;540 1.19 1.00

4 350 1.14 0.98

1 490;560 0.75 1.00

12 hrs 2 500;820 1.14 0.93 >820 820

3 300 1.05 0.96

4 530 0.87 0.99

- 100 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 170;180 0.99 1.04

2 hrs 2 150 1.04 1.02 >180 180

3 46 0.97 1.06

4 29 0.85 0.98

1 120;260;280 0.98 1.05

4 hrs 2 170 0.86 0.91 >300 300

3 230 0.90 0.96

4 300 0.94 0.97

1 340;350;440 0.94 0.92

6 hrs 2 210;230 1.03 1.05 >440 440

3 210 0.98 0.96

4 310 1.04 1.00

1 340;410;430 1.00 1.01

8 hrs 2 600 0.87 0.99 >600 600

3 350 0.90 1.00

4 390 1.02 0.94

1 570;700;770 0.63 0.88

10 hrs 2 560 0.95 0.89 770 830

3 610;830 0.85 0.92

4 710 1.19 0.93

1 1100;1200 1.11 0.88

12 hrs 2 430;780 1.01 0.93 >1200 1200

3 780 0.84 0.91

4 980 1.05 0.89

- 101 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 310;460;800# 0.37 1.00

12 hrs 2 760 1.04 0.99 800 760

3 510 0.97 0.97

4 730 0.93 0.98

1 810;1000* 0.00 0.94

18 hrs 2 430 0.89 0.97 1000 700

3 700 0.95 0.97

4 730;1300# 0.18 0.94

1 980;1400;1600# 0.27 0.97

24 hrs 2 1600 0.70 0.94 1100 1600

3 1000 0.89 0.96

4 990;1100# 0.19 0.93

1 890;1100;1100* 0.00 0.91

3 days 2 1000# 0.18 0.95 1000 740

3 620 0.94 1.01

4 740 1.28 0.97

1 1300;1400;1600* 0.00 0.94

7 days 2 1700# 0.23 0.92 1600 1100

3 1100 0.94 0.96

4 950 1.15 1.00

1 1500;1800;1900* 0.00 0.97

28 days 2 1400* 0.00 0.97 1400 1200

3 1200 0.89 0.96

4 1100;1500 0.23 0.95

- 102 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 590;800;1100# 0.18 0.88

12 hrs 2 1600 1.06 0.96 1100 1600

3 810 1.01 0.93

4 1900* 0.00 0.85

1 990 0.80 0.93

18 hrs 2 -* 0.00 0.91 1200 990

3 900;1200# 0.18 0.87

4 690 0.86 0.95

1 1500;1600* 0.00 0.99

24 hrs 2 1500 0.89 0.97 1500 1500

3 940 0.74 0.97

4 1400;1500# 0.10 0.97

1 1300;1400* 0.00 0.98

3 days 2 1000;1500* 0.00 0.96 1400 1100

3 980 1.03 0.99

4 1100 0.95 0.92

1 1500;1700# 0.21 0.93

7 days 2 1500;1600* 0.00 0.87 1600 720

3 720 0.84 0.97

4 1700 0.59 0.93

1 1100;1500;1800* 0.00 0.94

28 days 2 1400# 0.19 0.96 1400 <1400

3 1500# 0.33 0.97

4 1900* 0.00 0.97

- 103 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 22;90 1.07 0.98

2 hrs 2 22 0.89 0.94 >90 90

3 30 1.01 1.01

4 32 0.76 0.91

1 230;440 0.95 0.96

4 hrs 2 490 0.99 0.98 >620 620

3 280 1.05 1.02

4 620 0.95 0.98

1 200 0.96 1.00

6 hrs 2 84;190;260 1.26 1.03 >410 410

3 290 1.04 0.97

4 410 1.11 1.00

1 210;250;280 1.11 1.04

8 hrs 2 380 0.94 1.03 >380 380

3 320 1.04 0.97

4 400;480 ? 1.04

1 340;400;540 0.95 0.97

10 hrs 2 620 0.89 0.95 >790 790

3 460 1.06 1.03

4 650;790 0.88 0.99

1 260;480;560 0.95 0.99

12 hrs 2 370 1.05 0.99 >860 860

3 470 0.87 0.93

4 860 0.97 0.98

- 104 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 20;47 1.03 1.05

2 hrs 2 36 1.13 1.11 >69 69

3 29 1.18 1.03

4 69 1.05 1.07

1 180;240;350 1.13 1.04

4 hrs 2 460 1.08 1.01 >460 460

3 240 1.01 0.98

4 370 1.08 1.02

1 410;490;510 0.83 0.97

6 hrs 2 670 0.82 0.99 660 670

3 630 1.05 1.01

4 660 0.65 0.97

1 520;960;1100 1.23 1.04

8 hrs 2 860 1.06 0.96 860 1100

3 790 0.96 0.98

4 860 0.42 1.01

1 440;640;660 1.11 1.03

10 hrs 2 760 1.00 1.01 >760 760

3 450 0.94 0.99

4 730 1.04 0.99

1 550;560;830 1.03 0.88

12 hrs 2 580 1.06 1.00 >830 830

3 500 1.04 0.96

4 670 1.15 0.91

- 105 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 550;610;1100# 0.12 0.94

12 hrs 2 680 0.91 0.95 1100 1000

3 580 0.95 1.02

4 1000 0.83 0.96

1 750;850# 0.25 1.08

18 hrs 2 760 1.09 1.05 850 1000

3 1000 0.63 0.98

4 1000 1.12 1.01

1 870;1300# 0.25 0.99

24 hrs 2 1300 0.87 0.94 1300 1300

3 890 0.90 0.96

4 1500* 0.00 0.94

1 1100;1300;2000* 0.00 0.93

3 days 2 1400 1.09 0.88 1700 1400

3 940 0.89 0.99

4 1700* 0.00 0.98

1 1300;1400* 0.00 0.97

7 days 2 1200# 0.21 0.94 1200 810

3 810 0.91 0.99

4 1400;1500;1700* 0.00 0.99

1 1600;1900;1900 0.78 0.98

28 days 2 1400;1600;1700# 0.00 0.97 1500 1900

3 1400 0.89 0.96

4 1500# 0.35 0.96

- 106 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 680;700 0.78 0.92

12 hrs 2 430;640* 0.00 0.88 640 700

3 450 0.94 0.96

4 410 0.81 0.91

1 910;1100;1400# 0.21 0.99

18 hrs 2 980 0.58 1.05 980 780

3 780 0.96 1.02

4 1300# 0.19 1.02

1 1300;1600;1700# 0.39 1.01

24 hrs 2 1200 0.61 0.98 1200 1300

3 1300 1.01 1.03

4 1700# 0.20 0.98

1 1300;1400# 0.19 0.86

3 days 2 1700* 0.00 0.82 1400 900

3 900 0.81 1.01

4 1500;1600# 0.19 1.11

1 1300;1400;2000# 0.31 0.99

7 days 2 1300 0.22 0.92 1300 1300

3 1300 0.91 0.96

4 1500;1600* 0.00 0.98

1 1700;2000;2200* 0.00 1.02

28 days 2 1600 0.43 1.01 1400 980

3 1400* 0.00 1.00

4 980 0.95 1.00

- 107 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 49;98;110 0.83 1.00

2 hrs 2 38;45;53 0.89 1.04 >130 130

3 28;88 0.77 1.06

4 100;130 0.83 1.09

1 74;80 1.04 1.05

4 hrs 2 50;110 0.89 1.03 190 270

3 74;98;270 0.98 1.10

4 150;180;190 0.61 1.04

1 290;350;420 1.22 1.10

6 hrs 2 - 0.84 1.06 >420 420

3 100;110;130 0.90 0.95

4 110 1.10 1.01

1 130;150;330 0.99 1.04

8 hrs 2 150;210;220 1.04 1.00 >450 450

3 160;180;180 0.98 1.07

4 430;450 0.99 1.01

1 230;290;300 1.15 1.06

10 hrs 2 210;250 1.19 1.17 >390 390

3 170;380 1.20 1.25

4 240;390 1.13 1.25

1 210;210;240 1.02 1.05

12 hrs 2 230;390 0.95 1.04 >420 420

3 200;200;210 1.11 1.02

4 320;420 0.82 1.03

- 108 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 55;68;97 0.79 0.96

2 hrs 2 91 1.09 0.99 >97 97

3 27;42 0.90 0.99

4 61 0.92 0.89

1 74;100 0.87 0.99

4 hrs 2 45;95 1.00 1.13 >110 110

3 110 0.94 1.12

4 94;110 0.97 1.18

1 82;160;190 1.05 0.83

6 hrs 2 45 0.87 0.90 >190 190

3 96;150 0.81 0.88

4 87;140 1.09 0.85

1 100;250 0.71 1.01

8 hrs 2 180 0.85 1.08 >250 250

3 42;110 0.72 0.94

4 41;200 1.00 1.12

1 180;260 0.99 0.90

10 hrs 2 160 1.00 0.91 190 260

3 190 0.46 1.00

4 200 0.82 0.88

1 170;290;320 0.72 1.12

12 hrs 2 370 0.79 1.03 >390 390

3 250;390 0.82 1.00

4 240 0.97 1.09

- 109 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 380;430;520 0.60 0.94

12 hrs 2 1500* 0.00 0.92 520 <520

3 680;880# 0.35 0.98

4 580;970;1100 0.40 1.00

1 840;930;1300 0.81 1.02

18 hrs 2 420;560;890 0.91 1.01 1000 1300

3 870;1000* 0.00 1.03

4 320 0.78 1.01

1 1400;1900 0.49 1.02

24 hrs 2 1300 0.57 1.04 1300 720

3 2100* 0.00 1.00

4 720 0.81 1.01

1 270;590;610* 0.00 1.03

3 days 2 230;400;500 0.86 1.01 610 500

3 350;710* 0.00 1.02

4 210;300 0.81 1.04

1 240 0.90 0.98

7 days 2 940# 0.52 1.06 940 910

3 500;1100;1400* 0.00 0.89

4 910 1.01 1.01

1 -* 0.00 1.00

28 days 2 - 0.85 1.00 ? ?

3 -* 0.00 0.95

4 - 0.58 0.99

- 110 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 370;380;500* 0.00 0.96

12 hrs 2 380* 0.00 0.96 380 510

3 510 0.72 0.94

4 280 0.86 1.01

1 490;550* 0.00 0.96

18 hrs 2 330;560* 0.00 0.97 330 180

3 330 0.68 0.91

4 180 0.81 0.95

1 750;920# 0.42 0.93

24 hrs 2 1300* 0.00 1.07 920 1200

3 1200 0.72 0.99

4 1100 1.00 1.01

1 - 0.92 1.04

3 days 2 250;850;1800* 0.00 0.89 1800 <1800

3 - 0.92 0.99

4 -* 0.00 0.96

1 800;990;1000* 0.00 0.94

7 days 2 1100;1600# 0.00 0.86 1000 690

3 690 0.95 0.96

4 2200* 0.00 0.95

1 570;830 0.87 0.99

28 days 2 790* 0.00 1.13 790 830

3 -* 0.00 1.12

4 - 0.97 1.17

- 111 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 76;78 0.99 1.02

2 hrs 2 110;200;320 1.18 1.02 74 320

3 74 0.68 1.03

4 79 1.00 1.02

1 63;70 1.30 1.03

4 hrs 2 73;120 0.18 1.06 52 70

3 52 0.46 1.04

4 50 0.89 1.04

1 60;67 0.94 1.06

6 hrs 2 170;200 1.18 1.06 >200 200

3 96 0.78 1.09

4 120 0.96 1.07

1 81;81 1.08 1.04

8 hrs 2 62;79 0.74 1.09 >180 180

3 150 0.86 0.98

4 180 0.89 1.04

1 81;120 0.91 1.02

10 hrs 2 150;170 1.02 1.00 >170 170

3 110 1.01 0.99

4 120 0.97 1.00

1 170;170 0.98 0.96

12 hrs 2 160;160 0.89 0.98 350 260

3 260 0.88 1.01

4 350# 0.00 1.10

- 112 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 50;88 1.11 0.98

2 hrs 2 46 0.63 1.03 46 88

3 30 0.84 1.00

4 35 0.76 1.00

1 170;250 1.04 1.10

4 hrs 2 69 1.15 1.09 >250 250

3 95 0.71 0.93

4 210 0.99 1.04

1 74;150 1.23 0.96

6 hrs 2 130;140 1.08 0.94 >200 200

3 200 0.96 0.96

4 200 1.09 0.99

1 210 1.01 1.04

8 hrs 2 170 0.72 1.00 >230 230

3 220 0.92 1.02

4 230 0.99 1.01

1 390;420 1.01 1.07

10 hrs 2 530 0.56 0.98 530 530

3 530 0.84 1.04

4 340 1.10 1.02

1 410;550 0.99 1.00

12 hrs 2 550 0.86 1.05 >700 700

3 700 0.88 1.00

4 240 0.74 1.03

- 113 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 110;300;520 0.80 1.03

12 hrs 2 180;180;400 0.68 0.98 250 520

3 73 0.89 1.02

4 250 0.68 0.98

1 220;250;320* 0.00 0.97

18 hrs 2 480 0.85 0.95 310 510

3 310# 0.25 0.95

4 510 0.89 0.96

1 410 0.99 1.06

24 hrs 2 230;360* 0.00 1.01 360 620

3 550;570# 0.46 0.97

4 560;620 1.22 0.92

1 1300 0.72 0.96

3 days 2 690;750;920 0.77 0.99 1100 1300

3 750;840;1100# 0.18 0.93

4 860 0.76 1.00

1 830;1000;1300* 0.00 0.96

7 days 2 810;1200 0.27 0.94 1200 <1200

3 1200 0.32 0.98

4 1800* 0.00 0.96

1 2100* 0.00 0.90

28 days 2 1500 0.33 0.89 1400 750

3 1400* 0.00 0.92

4 750 0.88 0.90

- 114 -


Age athammer

test

Prismno.


applied(mm/s)

Tensilestrength

ratio,R1

Compressivestrength

ratio,R2

ppv1(mm/s)

ppv2(mm/s)

1 590;600 0.74 1.02

12 hrs 2 560 0.90 0.94 480 600

3 480# 0.40 0.97

4 580 0.76 0.98

1 810;960;1000# 0.32 1.12

18 hrs 2 850;970;1100 0.50 1.05 1000 1000

3 1000 1.07 1.05

4 1000 1.07 1.13

1 1100;1100# 0.17 0.99

24 hrs 2 860# 0.30 0.87 860 990

3 990 0.93 0.94

4 1000# 0.57 0.89

1 1100;1800* 0.00 0.97

3 days 2 1700# 0.26 1.04 1400 <1400

3 1400 0.50 1.00

4 -* 0.00 1.00

1 1600;1700# 0.22 1.05

7 days 2 1600 0.55 1.02 1400 <1400

3 1500* 0.00 1.05

4 1400# 0.04 1.08

1 1400;2000;2000* 0.00 1.08

28 days 2 2000 0.89 1.07 1600 2000

3 1400 1.03 1.06

4 770;1600* 0.00 1.05

- 115 -

5.2 Lower Bound Estimates of Vibration Resistance

As can be seen from Tables 5.1 - 5.24, there does not appear to be any simple relationbetween the intensity of the shock vibration applied and the corresponding values of R1 andR2 . Moreover, the value of ppv2 is sometimes higher than that of ppv1, albeit theoreticallyppv2 should be lower than ppv1. Hence, the vibration resistances of the concrete mixestested cannot be precisely determined.

An attempt of applying theoretical analysis to the ppv results will be made in the nextchapter, but it will soon be seen that any rigorous analysis is extremely difficult. In themeantime, it should be more prudent to find out the lower bound of the vibration resistancesof the concrete mixes tested and make use of the lower bound estimates to set safe ppv limitsfor preventing vibration damages to concrete.

From each pair of ppv1 and ppv2 values, a lower bound estimate of the vibrationresistance of the concrete may be obtained as:

ppv3 = minimum (ppv1, ppv2) …(5.5)

In cases where none of the hammered prisms in the same set had failed (in these cases, ppv1is given a value equal to “>ppv2”), ppv3 is taken to be equal to “ppv2”. On the other hand,in cases where all the hammered prisms had failed (in these cases, ppv2 is given a value equalto “<ppv1”), ppv3 is taken to have a value equal to “<ppv1” because in this situation, thevalue of ppv1 is not really a lower bound of the vibration resistance of the concrete tested(there was the possibility that the concrete tested could have failed at a somewhat lowervibration intensity than ppv1). Therefore, when the value of ppv3 is given a “<” sign, careshould be taken when interpreting its implications that this is not really a lower bound value.

The ppv3 values of the concrete mixes tested are listed in Tables 5.25 - 5.27 for ages ofless than 12 hrs. and in Tables 5.28 - 5.30 for ages of 12 hrs. to 28 days. Again, the ppv3values demonstrate that the vibration resistance of concrete is very erratic. Taking theresults presented in Table 5.25 as an example, it can be seen that for similar concretes (all theresults presented are those of grade 20 concretes cured at 20ºC - 30ºC), the ppv3 values at thesame age of 4 hrs. can range from 79 mm/s to 700 mm/s and that even for the same concrete,e.g. G20/25PFA/20C, the ppv3 value can change from a value of 1700 mm/s at 8 hrs. to520 mm/s at 10 hrs.

No obvious effects of using PFA up to 25% and curing temperature within the range of20ºC - 30ºC on the vibration resistance of concrete can be identified from the ppv3 resultspresented in the tables. Any probable effects of PFA and curing temperature within theranges studied have been over-shadowed by the large intrinsic variation of the vibrationresistance of concrete. In view of this situation, it is suggested that when setting vibrationcontrol limits to avoid vibration damages to concrete, a single set of vibration limits thatapplies to concretes with or without PFA and cured within the temperature range of 20ºC -30ºC should be used; it does not seem to be worthwhile to have more refined vibration controllimits that take these two factors into account. For the purpose of establishing a single set ofvibration control limits for a specific grade of concrete, the lower bounds of the ppv3 valuesof the four batches of concrete in the same grade are worked out and tabulated in the last rowof each table in Tables 5.25 - 5.30. These lower bounds should apply to all concretes of the

- 116 -

specified grade regardless of whether they contain PFA and their curing temperature providedtheir PFA content and curing temperature fall within the ranges studied.

The lower bound estimates of the vibration resistances of G20, G30 and G40 concretesare plotted against time in Figs. 5.1 and 5.2 to illustrate how they vary with the age of theconcrete. It is seen that the lower bound estimates of the vibration resistance of concretegenerally start at about 50 mm/s while the concrete is still fresh, increase to about 300 mm/sat 12 hrs. when the concrete is already hardened but is still relatively weak, and then furtherincrease to more than 700 mm/s when the concrete has fully developed its strength. Theincrease of these lower bound values with time is, however, not monotonic because they aresuperimposed with large intrinsic variation of the vibration resistance of concrete.

Table 5.25 - Lower bound of vibration resistances of G20 concretes - Series 1 tests

Batch no. ppv3 at various ages (mm/s)

2 hrs 4 hrs 6 hrs 8 hrs 10 hrs 12 hrs

G20/0PFA/20C 330 380 290 420 480 410

G20/0PFA/30C 550 79 180 180 820 860

G20/25PFA/20C 67 700 130 1700 520 660

G20/25PFA/30C 56 150 310 1100 310 310

lower bound forG20 concretes 56 79 130 180 310 310




G30/0PFA/20C 100 500 670 490 410 820

G30/0PFA/30C 180 300 440 600 770 1200

G30/25PFA/20C 90 620 410 380 790 860

G30/25PFA/30C 69 460 660 860 760 830


- 117 -




G40/0PFA/20C 130 190 420 450 390 420

G40/0PFA/30C 97 110 190 250 190 390

G40/25PFA/20C 74 52 200 180 170 260

G40/25PFA/30C 46 250 200 230 530 700




12 hrs 18 hrs 24 hrs 3 days 7 days 28 days

G20/0PFA/20C 570 890 350 1200 1200 2300

G20/0PFA/30C 910 990 840 2500 2900 2900

G20/25PFA/20C <340 1200 620 1400 1000 2300

G20/25PFA/30C 700 880 1700 <1200 2300 <2900

lower bound forG20 concretes <340 880 350 <1200 1000 2300




G30/0PFA/20C 760 700 1100 740 1100 1200

G30/0PFA/30C 1100 990 1500 1100 720 <1400

G30/25PFA/20C 1000 850 1300 1400 810 1500

G30/25PFA/30C 640 780 1200 900 1300 980


- 118 -




G40/0PFA/20C <520 1000 720 500 910 ?

G40/0PFA/30C 380 180 920 <1800 690 790

G40/25PFA/20C 250 310 360 1100 <1200 750

G40/25PFA/30C 480 1000 860 <1400 <1400 1600


5.3 Recommended Vibration Control Limits

Looking at the lower bound estimates plotted in Figs. 5.1 and 5.2, it is found verydifficult to set different vibration control limits for different grades of concrete, although it isbelieved that the grade of the concrete should have a bearing on the vibration resistance. Toovercome this difficulty, a lower bound for all the concrete mixes tested, i.e. concretes ofgrade 20 to 40, without PFA or with PFA up to 25%, and cured at a temperature within 20ºC -30ºC, is worked out and used to establish a single set of vibration control limits for all theseconcretes. The lower bound value for all the concrete mixes tested is just the smallest valueof ppv3 at a particular age among the twelve batches of concrete tested. These lower boundvalues of vibration resistance for all the concrete mixes tested are listed together with thelower bound estimates for the different grades of concrete in Tables 5.31 and 5.32. It isnoteworthy that whilst the “<” sign appears in some of the lower bound estimates for anindividual grade of concrete meaning that the values with the sign attached are not reallylower bounds, the “<” sign no longer appears in the lower bound values for all the concretemixes tested indicating that all the lower bound values arrived at are true lower bound values.

For easier visualization, the lower bound values of vibration resistance at various agesfor all the concrete mixes tested are plotted against time in Figs. 5.3 and 5.4. From these twographs, it is seen that the lower bound value of vibration resistance for all the concrete mixestested changes more smoothly with the age of the concrete and is less erratic compared to thelower bound estimates for the individual grades of concrete because of the larger number oftest samples from which the lower bound for all concrete mixes tested is derived.

Before a set of safe ppv limits can be established, there is still the problem of fixing anappropriate value for the factor of safety to be used. Although the lower bound values ofvibration resistance listed in Tables 5.31 and 5.32 which are derived from a fairly largenumber of vibration tests should be reasonably reliable, it is still considered advisable to adopta cautious approach of using a relatively large value for the factor of safety. Bearing in mindthat the effects of shock vibration on concrete are quite erratic and not yet fully understood, avalue of 5 is recommended.

- 119 -

Dividing the lower bound values of vibration resistance listed in Tables 5.31 and 5.32by a factor of safety of 5 and adjusting the resulting values slightly so that they map into asmooth and monotonic function of the age of concrete, the following set of safe ppv limits isobtained:

Age of concrete Safe ppv limit (mm/s)≤ 4 hrs. 10

6 - 8 hrs. 2010 - 12 hrs. 30

18 hrs. 4024 hrs. 703 days 1007 days 125

≥ 28 days 150

The above recommended vibration control limits are also given in the last rows of Tables 5.31and 5.32 for easy comparison with the lower bound values of vibration resistance. Sincethese limits are derived from the lower bound values for all the concrete mixes tested, theyshould be applicable to all concretes whose grade, PFA content and curing temperature fallwithin the ranges studied.

These vibration control limits are compared to those adopted by others in Figs. 5.5 to5.8. It is seen that the recommended vibration limits for green concrete (concrete at age lessthan 24 hrs.) are generally higher than those adopted by Gamble and Simpson (1985) andBryson and Cooley (1985) but lower than those of Olofesson (1988) for concrete at age lessthan 4 hrs. For concrete more than 1 day old, the above limits are higher than all thoseadopted by Gamble and Simpson (1985), Bryson and Cooley (1985) and Olofesson (1988).However, the recommended limits are lower than those proposed by Hulshizer and Desai(1984) at virtually all ages.

- 120 -

Table 5.31 - Lower bound of vibration resistances of all concretes tested - Series 1 tests

Batch no. vibration resistance at various ages (mm/s)





lower boundfor all

concretes tested46 52 130 180 170 260

recommendedvibration

control limits10 10 20 20 30 30

Table 5.32 - Lower bound of vibration resistances of all concretes tested - Series 2 tests

Batch no. vibration resistance at various ages (mm/s)


lower bound forG20 concretes <340 880 350 <1200 1000 2300



lower boundfor all

concretes tested250 180 350 500 690 750

recommendedvibration

control limits30 40 70 100 125 150

- 125 -

6. THEORETICAL ANALYSIS

6.1 Mean Value Estimates of Vibration Resistance

In the previous chapter, the lower bounds of the vibration resistances of the concretemixes tested are used to establish vibration control limits to ensure that the vibration limitsarrived at are on the safe side. However, for theoretical analysis, mean value estimates,which may not be on the safe side but should be more accurate, seem more appropriate.Mean value estimates of vibration resistance, abbreviated as ppv4, may be obtained as follows:

ppv4 =12

(ppv1 + ppv2) …(6.1)

In cases where none of the hammered prisms in the same set had failed (in these cases, ppv1is given a value equal to “>ppv2”), ppv4 is taken to be equal to “>ppv2” meaning that theactual vibration resistance of the concrete tested should be higher than ppv2. On the otherhand, in cases where all the hammered prisms had failed (in these cases, ppv2 is given a valueequal to “<ppv1”), ppv3 is taken to have a value equal to “<ppv1” meaning that the actualvibration resistance of the concrete tested should be lower than ppv1. Values of ppv4without any “<” or “>” sign should be more reliable as the ppv1 and ppv2 values from whichthey are derived are true ppv values at which failure of the concrete had or had not occurred.Values of ppv4 with either a “<” or “>” sign give only upper and lower limits respectively ofthe vibration resistance.

The ppv4 values so obtained are tabulated together with the material properties of theconcrete mixes tested in Tables 6.1 - 6.3 for concretes at age ≤ 12 hrs. and in Tables 6.4 - 6.6for concretes at age ≥ 12 hrs. Detailed analysis of the ppv4 results and correlation of theppv4 values to the material properties of the concrete are presented in the next two sections,where an attempt of identifying the most important material parameter that determines thevibration resistance is also made.

- 126 -

Table 6.1 - Vibration resistances of G20 concretes at age ≤ 12 hrs.

Batch no. Age athammer test

Penetrationresistance

(MPa)

ppv4(mm/s)

2 hrs 0.01 >330

4 hrs 1.43 >380

G20/0PFA/20C 6 hrs - 295

8 hrs. - >420

10 hrs. - >480

12 hrs. - >410

2 hrs 0.03 >550

4 hrs 2.97 120

G20/0PFA/30C 6 hrs - >180

8 hrs. - >180

10 hrs. - >820

12 hrs. - >860

2 hrs 0.01 >67

4 hrs 0.22 >700

G20/25PFA/20C 6 hrs 1.74 200

8 hrs. - 1700

10 hrs. - 1310

12 hrs. - 2030

2 hrs 0.01 >56

4 hrs 0.19 >150

G20/25PFA/30C 6 hrs 2.67 >310

8 hrs. - >1100

10 hrs. - 360

12 hrs. - 355

- 127 -




(MPa)

ppv4(mm/s)

2 hrs 0.09 >100

4 hrs 1.03 >500

G30/0PFA/20C 6 hrs - >670

8 hrs. - 495

10 hrs. - 475

12 hrs. - >820

2 hrs 0.11 >180

4 hrs 3.54 >300

G30/0PFA/30C 6 hrs - >440

8 hrs. - >600

10 hrs. - 800

12 hrs. - >1200

2 hrs 0.03 >90

4 hrs 0.26 >620

G30/25PFA/20C 6 hrs 2.58 >410

8 hrs. - >380

10 hrs. - >790

12 hrs. - >860

2 hrs 0.04 >69

4 hrs 1.18 >460

G30/25PFA/30C 6 hrs - 665

8 hrs. - 980

10 hrs. - >760

12 hrs. - >830

- 128 -




(MPa)

ppv4(mm/s)

2 hrs 0.03 >130

4 hrs 0.43 230

G40/0PFA/20C 6 hrs 3.40 >420

8 hrs. - >450

10 hrs. - >390

12 hrs. - >420

2 hrs 0.15 >97

4 hrs 1.53 >110

G40/0PFA/30C 6 hrs - >190

8 hrs. - >250

10 hrs. - 225

12 hrs. - >390

2 hrs 0.01 197

4 hrs 0.28 61

G40/25PFA/20C 6 hrs 2.20 >200

8 hrs. - >180

10 hrs. - >170

12 hrs. - 305

2 hrs 0.01 67

4 hrs 0.32 >250

G40/25PFA/30C 6 hrs 4.19 >200

8 hrs. - >230

10 hrs. - 530

12 hrs. - >700

- 129 -

Table 6.4 - Vibration resistances of G20 concretes at age ≥ 12 hrs.

Batch no.Age at

hammertest

f cu

(MPa)f t

(MPa)Ed

(GPa)c

(m/s)ppv4

(mm/s)

12 hrs 3.9 0.43 - - 605

18 hrs 5.8 0.74 19.4 3140 905

G20/0PFA/20C 24 hrs 8.9 1.00 22.9 3410 525

3 days 12.7 1.31 26.9 3700 1350

7 days 17.7 1.54 30.6 3940 1400

28 days 31.9 2.46 34.7 4200 2300

12 hrs 4.8 0.48 15.2 2780 1055

18 hrs 6.3 0.81 19.5 3150 1145

G20/0PFA/30C 24 hrs 11.0 1.21 22.5 3380 1020

3 days 14.7 1.56 23.6 3460 2600

7 days 21.5 1.99 28.1 3780 3000

28 days 32.3 2.68 35.1 4220 3150

12 hrs 1.3 0.13 8.4 2070 <340

18 hrs 3.1 0.28 14.3 2700 1750

G20/25PFA/20C 24 hrs 4.9 0.49 16.4 2890 685

3 days 8.9 1.01 21.6 3310 1850

7 days 10.5 1.37 25.9 3630 1800

28 days 22.2 1.76 31.8 4020 2700

12 hrs 3.0 0.28 7.8 1990 >700

18 hrs 3.4 0.32 12.9 2560 905

G20/25PFA/30C 24 hrs 5.0 0.61 16.8 2920 >1700

3 days 9.9 1.09 21.6 3310 <1200

7 days 12.2 1.08 26.8 3690 2900

28 days 24.5 2.18 31.3 3990 <2900

- 130 -


Batch no.Age at

hammertest

f cu

(MPa)f t

(MPa)Ed

(GPa)c

(m/s)ppv4

(mm/s)

12 hrs 4.1 0.40 13.9 2640 780

18 hrs 7.5 0.76 18.9 3080 850

G30/0PFA/20C 24 hrs 10.2 0.94 20.4 3200 1350

3 days 18.6 1.82 28.6 3790 870

7 days 28.7 2.09 32.1 4010 1350

28 days 39.3 2.87 36.5 4280 1300

12 hrs 5.9 0.52 16.3 2860 1350

18 hrs 9.0 0.81 20.2 3180 1095

G30/0PFA/30C 24 hrs 10.4 0.99 20.9 3240 1500

3 days 21.9 1.79 27.7 3730 1250

7 days 37.1 2.72 33.8 4120 1160

28 days 41.9 3.07 35.9 4240 <1400

12 hrs 2.6 0.27 12.2 2490 1050

18 hrs 4.8 0.41 15.9 2840 925

G30/25PFA/20C 24 hrs 7.4 0.75 19.0 3110 1300

3 days 10.4 1.09 24.7 3540 1550

7 days 18.9 1.62 29.3 3860 1005

28 days 33.9 2.54 34.9 4210 1700

12 hrs 3.6 0.37 12.6 2530 670

18 hrs 7.2 0.62 18.8 3090 880

G30/25PFA/30C 24 hrs 8.6 0.77 20.1 3200 1250

3 days 14.8 1.30 27.2 3720 1150

7 days 24.4 2.20 30.1 3910 1300

28 days 43.1 3.04 35.7 4260 1190

- 131 -


Batch no.Age at

hammertest

f cu

(MPa)f t

(MPa)Ed

(GPa)c

(m/s)ppv4

(mm/s)

12 hrs 2.0 0.29 10.3 2290 <520

18 hrs 5.6 0.66 21.5 3310 1150

G40/0PFA/20C 24 hrs 7.1 0.73 23.1 3430 1010

3 days 23.5 1.79 29.0 3840 555

7 days 31.2 2.40 32.6 4070 925

28 days 46.9 3.17 38.2 4410 ?

12 hrs 1.6 0.09 - - 445

18 hrs 13.0 1.06 22.3 3370 255

G40/0PFA/30C 24 hrs 13.4 1.12 23.0 3420 1060

3 days 27.0 2.08 28.4 3800 <1800

7 days 38.9 2.60 31.3 3990 845

28 days 50.8 3.26 36.0 4280 810

12 hrs 0.7 0.03 - - 385

18 hrs 2.0 0.33 8.8 2110 410

G40/25PFA/20C 24 hrs 3.9 0.64 16.1 2860 490

3 days 18.0 1.83 27.1 3710 1200

7 days 26.5 2.31 31.3 3990 <1200

28 days 40.5 2.65 32.9 4090 1075

12 hrs 2.2 0.20 8.1 2030 540

18 hrs 4.3 0.37 18.1 3030 1000

G40/25PFA/30C 24 hrs 8.0 0.65 19.8 3170 925

3 days 18.3 1.52 25.8 3620 <1400

7 days 28.0 1.96 29.6 3880 <1400

28 days 48.0 3.03 36.6 4310 1800

- 132 -

6.2 Vibration Resistance of Concrete at Age ≤ 12 Hours

At concrete ages of less than 12 hours, due to the difficulty of removing the moulds ofthe concrete for any stiffness or strength measurement, the only mechanical properties of theconcrete measured were the penetration resistance and stiffening time of the mortar fraction ofthe concrete. These material properties have been presented in Section 4.2. Herein, thepenetration resistances of the concretes at ages of 2 hrs., 4 hrs. and 6 hrs. are tabulatedalongside the ppv4 results in Tables 6.1 - 6.3 for detailed study of any correlation between thevibration resistance and the penetration resistance of the concrete.

The ppv4 results presented in Tables 6.1 - 6.3 are plotted against the age of theconcrete for the G20, G30 and G40 concretes in Figs. 6.1, 6.2 and 6.3 respectively. Whenthe ppv4 results are plotted in these figures, the “<” and “>” signs of the ppv4 values areignored (in fact, none of the ppv4 values has the “<” sign). From these figures, it is seen thatthe ppv4 results are very scattered but there is still a general trend that the vibration resistanceincreases with the age of the concrete. To better illustrate the spread of the test results, allthe data points for the different batches of concrete tested are plotted together in Fig.6.4.The large scatter of the ppv4 results reflects the erratic nature of the vibration resistance ofconcrete. Quite often, at a given concrete age, the range of the ppv4 values is severalhundred percent of the value of the lower bound of the range. In Fig.6.4, a best fit cubicpolynomial equation for the data points is also presented. This best fit polynomial equationis somehow very close to a linear equation as indicated by the straightness of the polynomialcurve. Hence, although the ppv4 results are very scattered due to the superposition on themof large intrinsic variation of the vibration resistance of concrete, their mean value increasessteadily and more or less linearly with the age of the concrete up to an age of about 12 hours.

In order to investigate whether the vibration resistance of concrete can be correlated tothe penetration resistance, the ppv4 values are plotted against the penetration resistance inFig.6.5 up to a penetration resistance of about 4.0 MPa - the maximum measuring capacity ofthe stiffening time tester used. Again, the data points plotted are very scattered. A best fitline for the data points is also drawn in the figure to reveal the trend of variation of thevibration resistance with the penetration resistance. From this line, it can be seen thatalthough the correlation between the vibration resistance and the penetration resistance israther weak, there is a marginal increase in the vibration resistance as the penetrationresistance increases from zero to about 4.0 MPa. Actually, at a penetration resistance of lessthan 4.0 MPa, the concrete is still plastic and thus behaves more like a colloidal material thana hardened solid. At this stage, the concrete has very little stiffness that could help to resistshock vibration. This is probably why the vibration resistance is almost independent of orincreases only very slightly with the penetration resistance. What seems more important isthe sensitivity of the mortar matrix of the plastic concrete to possible disruptions caused bythe shock vibration. For this same reason, the stiffening time of the concrete does not appearto have significant influence on the variation of the vibration resistance with time.

- 136 -

equations, exponential equation and power equation, have been tried and it turns out after aseries of regression analysis that a power equation gives the highest correlation index. Thebest fit power equation is found to be given by:

ppv4 = 585 26.0cu )(ƒ⋅ …(6.1)

in which the units for ppv4 and ƒcu are mm/s and MPa respectively. It is plotted in Fig.6.10to illustrate how it fits the data points. Due to the large scatter of the ppv4 results, the valueof R2 for this equation, though already the highest among those of the different forms ofequation tried, is only 0.24.

However, the Eqn.6.1 above gives only a mean value estimate of the vibrationresistance. As there might be large intrinsic variation in the vibration resistance, the actualvibration resistance of the concrete could be significantly higher or lower than the valuepredicted by this equation. What is more important is that there is a very high probability ofabout 50% of having the actual vibration resistance lower than the predicted mean value.Therefore, this equation is not considered safe to use in engineering practice. Like the use ofcharacteristic strength in structural design, it is proposed to use a characteristic value estimateof the vibration resistance instead of the above mean value estimate to cater for the relativelylarge intrinsic variation of vibration resistance. As for the case of characteristic strength, thecharacteristic vibration resistance is defined as the vibration level at which no more than 5%of the concrete is expected to fail. Assuming that the characteristic vibration resistance,denoted by ppvc, is also related to the compressive strength of the concrete by a powerequation of the form: ppvc = k(ƒcu)0.26, it may be estimated by finding the value of k such thatonly 5% of the data points fall below this curve. In Fig.6.10, there are a total of 60 datapoints. 5% of the number of data points is 3. The value of k such that only three datapoints fall below the curve and the third and fourth lowest data points are at equal distancefrom the curve is found by trial and error to be 290. Hence, the characteristic vibrationresistance may be estimated by the following equation:

ppvc = 290 26.0cu )(ƒ⋅ …(6.2)

This equation, plotted as a dotted line in Fig.6.10, may also be used to establish the vibrationcontrol limits for avoiding damage to concrete during shock vibration. Adopting a factor ofsafety of 5, the safe vibration limits may be obtained by dividing the ppvc values sodetermined by 5, as listed in Table 6.7. Bearing in mind that most concretes cured at anormal temperature of 20ºC - 30ºC develop a cube strength of at least 3 MPa in 24 hours andtheir full strength potential in 28 days, the safe ppv limits listed in Table 6.7 of 77 - 160 mm/sfor concretes with cube strength ranging from 3 to 50 MPa agree fairly well with thecorresponding safe ppv limits given in Section 5.3 of 70 - 150 mm/s for concretes at agebetween 24 hours to 28 days. Hence, in most cases, similar vibration control limits would beobtained regardless of whether the ppv limits based on age as given in Section 5.3 or the ppvlimits based on cube compressive strength as listed in Table 6.7 are used.

- 137 -

Table 6.7 - Vibration control limits based on mean cube compressive strength

ƒcu (MPa) ppvc (mm/s) Safe ppv limit (mm/s)

3.0 385 77

5.0 440 88

10.0 530 106

20.0 630 126

30.0 700 140

40.0 755 151

50.0 800 160

As the most likely damage that would be caused by shock vibration is cracking, thetensile strength of the concrete should have a more direct influence on the vibration resistancethan the compressive strength. In order to study how the tensile strength affected thevibration resistance, the ppv4 results are plotted against the split cylinder tensile strength ofthe concrete at the time of hammering in Fig.6.11. Again, the data points plotted on thegraph are fairly scattered. As before, several different types of curve: first and second orderpolynomial equations, exponential equation and power equation, have been tried to fit the datapoints and a power equation is found to give the best correlation. The best fit powerequation is given by:

ppv4 = 1098 ⋅ ( ) .f t0 29 …(6.3)

in which the units for ppv4 and f t are mm/s and MPa respectively. It is plotted in Fig.6.11to illustrate how it fits the data points. The value of R2 for this equation is 0.26, which isslightly better than the value of 0.24 for Eqn.6.1. Hence, it may be said that the vibrationresistance correlates to the tensile strength slightly better than to the compressive strength.

Following the method used previously for estimating the characteristic vibrationresistance from the cube compressive strength, the characteristic vibration resistance may alsobe estimated from the tensile strength by assuming that it is related to the tensile strength inthe form ppvc = k fcu( ) .0 29 and finding the value of k such that only 5% of the data points fallbelow the curve. The value of k so determined is found to be 527. Hence, thecharacteristic vibration resistance may be estimated from the tensile strength by the followingequation:

ppvc = 527 ⋅ ( ) .f t0 29 …(6.4)

The above equation is plotted as a dotted line in Fig.6.11. Dividing the ppvc values by afactor of safety of 5, a set of vibration control limits can be obtained, as listed in Table 6.8.

- 138 -

These vibration control limits may be compared to those listed in Table 6.7 by converting thecube compressive strength values in Table 6.7 to tensile strength values using the simple rulethat the tensile strength of concrete is approximately equal to 8 - 10 % of the cubecompressive strength. After the conversion, it can be seen that the safe ppv limits given inTable 6.8 agree fairly well with those listed in Table 6.7.

Table 6.8 - Vibration control limits based on split cylinder tensile strength

f t (MPa) ppvc (mm/s) Safe ppv limit (mm/s)

0.1 270 54

0.3 370 74

0.5 430 86

1.0 525 105

2.0 645 129

3.0 725 145

4.0 790 158

Before the theoretical study continues, it should be noted from the above analysis thatthe vibration resistance is neither proportional to the compressive strength nor proportional tothe tensile strength. The believe of many engineers that the vibration resistance isproportional to the strength of concrete (however, it has never been clear whether the strengthshould be the compressive strength or the tensile strength) is unfounded.

Since the maximum dynamic tensile strain εd induced during shock vibration isdependent on the longitudinal wave propagation velocity c as given by the followingequation:

εd =vc

…(6.5)

where v is the peak particle velocity of the shock wave, it is expected that the longitudinalwave propagation velocity may also have significant influence on the vibration resistance ofconcrete. To study any possible dependence of the vibration resistance on the value of c ,the ppv4 results are plotted against the corresponding values of c at the time of hammeringin Fig.6.12. From this figure, it can be easily seen that in general, the vibration resistance ishigher when the longitudinal wave propagation velocity is on the high side. As before,several different types of curve: first and second order polynomial equations, exponentialequation and power equation, have been tried to fit the data points and a power equation isfound to give the best correlation. The best fit power equation is given by:

- 139 -

ppv4 = 0.047 ⋅ ( ) .c 1 24 …(6.6)

in which the units for ppv4 and c are mm/s and m/s respectively. It is plotted in Fig.6.12 asa solid line. Due to the large scatter of the ppv4 results, the value of R2 for this equation isonly 0.21. Assuming that the characteristic vibration resistance is related to the value of cby ppvc = k c( ) .1 24 and finding the value of k such that only 5% of the data points fall belowthis curve, the equation that correlates the characteristic vibration resistance to thelongitudinal wave propagation velocity is obtained as:

ppvc = 0.023 ⋅ ( ) .c 1 24 …(6.7)

It is plotted in Fig.6.12 as a dotted line. For most hardened concretes, the value of c rangesfrom about 2000 m/s at an age of 12 hours to about 4000 m/s upon development of fullstrength. Substituting these values of c into Eqn.6.7, the corresponding values ofcharacteristic vibration resistance are evaluated as 285 mm/s and 673 mm/s respectively.Dividing these values by a factor of safety of 5, the safe ppv limits for hardened concrete areobtained as 57 mm/s for young hardened concrete at an age of 12 hours and 135 mm/s forfully hardened concrete at an age of about 28 days. These safe ppv limits are in goodagreement with the corresponding values presented in Section 5.3.

Making use of Eqn.6.5, the dynamic tensile strain εd ' at which the concrete failedwhen subjected to shock vibration may be estimated as ppv4/ c . The dynamic tensile strainvalues so evaluated are plotted against the cube compressive strength in Fig.6.13 to study howthe dynamic tensile strain capacity varies with the compressive strength. Curve fitting of thedata points yields the following equation for the dynamic tensile strain capacity:

εd ' = 321 ⋅ ( )f cu0.015 …(6.8)

The value of R2 for this equation is only 0.001 indicating that the dependence of thedynamic tensile strain capacity on the compressive strength of the concrete is rather weak.Following the method used before, the characteristic dynamic tensile strain capacity εd ' isobtained as:

εd ' = 160 ⋅ ( )f cu0.015 …(6.9)

The above two equations are plotted in Fig.6.13 as solid and dotted lines respectively. Fromthese two curves, it can be seen that the dynamic tensile strain capacity increased onlymarginally with the compressive strength; in fact, it increased by only about 4% when thecube compressive strength increased from 3 MPa to 30 MPa. It is also noted that forhardened concrete, whose cube compressive strength is expected to be higher than 3 MPa, thecharacteristic value of dynamic tensile strain capacity is generally larger than 160 micro-strain.Hence, for simplicity, the characteristic dynamic tensile strain capacity of hardened concretemay be just taken as a constant value of 160 micro-strain regardless of the actual strength ofthe concrete. Dividing this value of characteristic dynamic tensile strain capacity by a factorof safety of 5, a safe dynamic tensile strain limit of 32 micro-strain is obtained.

- 140 -

The above study on dynamic tensile strain capacity is of particular importance tounderground concrete structures which are restrained to deform together with the surroundingsoil (restrained structures). Let the wave propagation velocity in the surrounding soil bedenoted by cs . The maximum dynamic strain ε s produced by the propagation of a shockwave of peak particle velocity v through the soil is given by:

ε s =vcs

…(6.10)

If the surrounding soil has a lower stiffness than the concrete and hence a lower wavepropagation velocity, i.e. c cs < , then the strain produced in the soil will be larger than thatproduced in an independent concrete structure (a concrete structure free to deformindependently of the surrounding soil). However, since the underground concrete structureis restrained to deform together with the surrounding soil, the strain produced in the concreteis forced to be equal to that produced in the soil. Therefore, when evaluating the maximumdynamic tensile strain εd produced in restrained concrete structures, the smaller of the valuesof c and cs should be used, as in the following equation:

εd =vc csmin.( , )

…(6.11)

The vibration control should thus be exercised such that the dynamic tensile strain evaluatedby the above equation is within the safe dynamic tensile strain limit. Working the other wayround, the vibration control limits for restrained concrete structures may be set as:

safe ppv limit = min.( , )c cs × safe dynamic tensile strain limit …(6.12)

When c cs < , as for hardened concrete structures buried in soft soil, the safe ppv limit will begoverned by the value of cs . On the other hand, when c cs > , as for green concrete castagainst soil or hardened concrete structures embedded in hard rock, the safe ppv limit will begoverned by the value of c , or in other words, the surrounding soil will have no effect on thevibration control limit.

In Chapter 2, it has been mentioned that the vibration resistance of concrete may beestimated from the wave propagation velocity c , dynamic tensile strength f td and dynamicmodulus Ed using the following equation:

vc

Ef

dtd' = ...(6.12)

The dynamic tensile strength had not been measured in this project. Nevertheless, assuggested by Raphael (1984), the dynamic tensile strength may be taken simply as 1.5 timesthe static tensile strength. In order to investigate whether the actual measured values ofvibration resistance agree with the above theoretical prediction, the ppv4 results are plottedagainst the corresponding values of v' in Fig.6.14. A straight line whose equation is givenby ppv4 = v' is also drawn on the graph. It is seen that all the data points are above this line.

- 146 -

7. DISCUSSIONS

7.1 Failure Criterion for Vibration Resistance

The vibration resistance of concrete is the vibration intensity at which the concretefails. However, the failure criterion has never been properly defined. In fact, the literaturereview carried out in the Literature Review Report No.1 of the present studied revealed thatdifferent researchers used quite different failure criteria.

Hulshizer and Desai (1984) considered the tensile strength of concrete unimportant forthe reason that reinforced concrete is basically designed as a “cracked section”. Therefore,in their study of the effects of blasting vibration on concrete, no effort was made to test thetensile strength of the concrete; they tested only the compressive, shear and bond strengths ofthe vibrated concrete and compared the results to those of un-vibrated concrete to evaluate thevibration resistance of the concrete. Restricting only to compressive, shear and bondstrengths, they found no significant change in these strength values despite vibrations ofintensity up to 250 mm/s had been applied. At the end, they concluded there was no sign toindicate that the vibrated concrete would not perform as if they were un-vibrated andrecommended relatively high vibration limits for blasting control.

On the other hand, as cited by Dowding (1985), Esteves (1978) defined failure as“fracture of concrete” when he performed his hammering tests. It is not clear what fracturemeans. Judging by common sense, fracture probably means complete cracking of a wholesection of the concrete such that it breaks into at least two pieces or the appearance of at leastone visible crack on the concrete surface. As the concrete prisms tested were still containedin moulds when they were subjected to hammering, it was unlikely that they would break intopieces. Hence, the most likely failure criterion adopted was the appearance of at least onevisible crack on the concrete surface. Since the tensile strength of the hammered prisms hadnot been measured, it was not known whether the vibration applied had any effect on thetensile strength. Nevertheless, Esteves found that although there is a period of greatersusceptibility to vibration cracking between 10 and 20 hours after casting, the thresholdvibration for cracks to appear during this period is as high as 150 mm/s.

In the present study, a more definite failure criterion as “complete cracking of a wholesection of the concrete such that it breaks into at least two pieces or having more than 30%reduction in tensile or compressive strength” is adopted. Since shock vibrations have littleeffect on the compressive strength, whether there has been significant reduction in tensilestrength actually governs. Moreover, as breakage of the concrete into pieces imply reductionof its tensile strength to zero, the failure criterion may be simplified as “having more than30% reduction in tensile strength”. The reason for adopting this failure criterion is that incertain cases, even without any signs of observable cracks formed on the surface, the tensilestrength of the vibrated concrete might have already been significantly affected due todisturbance to its mortar matrix or formation of internal cracks.

Referring back to Tables 5.1 - 5.24, it is seen that among the 288 concrete prismswhich had been subjected to hammering at age ≤ 12 hours (Series 1 tests), 27 had failed(cracked or tensile strength reduced by more than 30%) but only one of them (Prism No.4 ofG40/25PFA/20C tested at 12 hours) had cracked. At ages of 10 hours or less, none of thehammered prisms had any observable cracks formed on them, although quite a number of

- 147 -

them had their tensile strengths reduced by more than 30%. Hence, at early ages of concrete(up to 10 hours), the effects of shock vibration are not easily observable by naked eyes.When making the judgment of whether the shock vibration applied has caused adverse effectson the concrete, it is not reliable to just consider the occurrence of visible cracks on thesurface because there is the possibility that the concrete has been adversely affected withoutany signs of cracking.

It is, therefore, important when comparing the vibration resistance results of differentresearchers to take into account the actual failure criteria used by them. The failure criterionadopted by Hulshizer and Desai is the most lenient. As a results, they produced very highvibration resistance results. The failure criterion adopted by Esteves is also on the lenientside as any possible effect on the tensile strength was not considered. These are the reasonswhy the vibration resistances obtained in the present study are relatively low compared to theresults of Hulshizer and Desai at virtually all ages of concrete and also to the results ofEsteves at ages of less than 12 hours.

7.2 Vibration Resistance During First Few Hours

The largest discrepancy between the vibration control limits being adopted by differentengineers lies in the vibration control limit to be applied during the first few hours before theconcrete sets.

Some researchers (Hulshizer and Desai 1984) suggested that blasting vibrationsapplied to concrete before it sets are just like additional vibration for compaction andtherefore can only be beneficial rather than detrimental. Hence, they adopted a very highvibration control limit of 100 mm/s during the first 3 hours. As cited by Olofesson (1988),Ontario Hydro had used the same vibration control limit of 100 mm/s during the first 10 hoursfor concrete cured at 5oC and during the first 5 hours for concrete cured at 21oC.

Other researchers (Akins and Dixon 1979; Gamble and Simpson 1985), however,imposed severe restrictions during the entire period from 0 to 24 hours after casting. Akinsand Dixon argued that although re-vibration of concrete at 2 - 4 hours after initial placementin such a way that the concrete is returned to a plastic condition may be beneficial, it isunlikely that the transient vibration due to blasting can return the concrete to a plasticcondition and afford the advantages of re-vibration. Shock vibrations sufficient to disturbthe concrete matrix but less than sufficient to return the concrete to plastic state can beharmful. Added with the considerations that little is known about the effects of vibration ongreen concrete and autogenous healing cannot be relied on, they imposed a vibration limit of5.1 mm/s during the entire period of green concrete, i.e. the first 24 hours. On the other hand,Gamble and Simpson simply assumed that the vibration resistance of concrete is directlyproportional to the compressive strength and as the strength of concrete is very low before ithardens, imposed vibration control limits which are even more restrictive than thoserecommended by Akins and Dixon during the first 12 hours.

With the large amount of test results produced in the present study, a settlement to thishistorical conjecture or dilemma can now be made. The test results indicate that thevibration resistance of concrete is lowest during the first few hours but would graduallyincrease with age until the concrete is fully hardened. Hulshizer and Desai’s view that the

- 148 -

concrete should be able to withstand very large vibration before it sets and it is only after theconcrete has set that it becomes particularly susceptible to vibration damage is unfounded.The period that the concrete is particularly susceptible to vibration damage actually startsbefore the concrete sets. In any case, it is considered unwise to set a high vibration controllimit for concrete before it sets. The changeover of permitted vibration intensity from a veryhigh level while the concrete is still wet to a very low level after the concrete has set is tooabrupt. Since the setting time (to be more exact, the stiffening time) of the concrete dependson the mix composition and is quite sensitive to the curing temperature, it is not easy toprecisely determine the changeover time and as a result, there is the danger that the concretemight have entered the low vibration resistance stage earlier than expected.

Nevertheless, the measured vibration resistance of green concrete is generallysubstantially higher than that assumed by Akins and Dixon. Hence, Akins and Dixon were abit overly conservative. The blasting operation would become very expensive if theirvibration control limits are adopted. The vibration control limits proposed in the presentstudy, which are substantially higher than those adopted by Akins and Dixon, will allow morepower blasting near green concrete and hence faster and more economical construction.

Finally, the test results clearly show that the vibration resistance is neither directlyproportional to the compressive strength nor to the tensile strength (the material parametersthat really determine the vibration resistance will be discussed in the next section). Gambleand Simpson’s assumption that the vibration resistance is proportional to strength is incorrect.The very low vibration control limits proposed by them would virtually ban all kinds ofblasting operation near green concrete.

7.3 Material Parameters Determining Vibration Resistance

For concrete at age less than 12 hours, the only material properties measured were thepenetration resistance of the mortar fraction and the stiffening time. However, it is foundthat the vibration resistance is almost independent of the penetration resistance. Anypossible dependence of the vibration resistance on the penetration resistance has been coveredup by the large intrinsic variation of the vibration resistance. The stiffening time is alsofound to have little influence on the change of vibration resistance with time. Hence, nomaterial property that would directly or indirectly influence the vibration resistance ofconcrete at age less than 12 hours has been identified. The only parameter which thevibration resistance is found to be closely related to is the age of the concrete. Regressionanalysis of the test results reveals that on average, the vibration resistance increases steadilyand more or less linearly with the age of the concrete up to 12 hours.

For concrete with age more than 12 hours, the compressive strength, tensile strength,dynamic and static moduli, and longitudinal wave propagation velocity had been measured.Although theoretically the compressive strength should not have a direct influence on thevibration resistance, as most material properties including those that may affect the vibrationresistance are either directly or indirectly related to the compressive strength, there should bean indirect relation between the vibration resistance and the compressive strength.Regression analysis of the test data indicates that the vibration resistance is approximatelyproportional to ( ) .f cu

0 26 . Since the most likely damage to concrete by shock vibration iscracking or reduction in tensile strength, the tensile strength is expected to have a direct

- 149 -

influence on the vibration resistance. Correlation of the vibration resistance to the tensilestrength reveals that the vibration resistance may be taken as proportional to ( ) .f t

0 29 . Thevibration resistance is, therefore, neither directly proportional to the compressive strength norto the tensile strength, and it is thus believed that both the compressive strength and the tensilestrength are not the material properties that would directly determine the vibration resistance.

On the other hand, regression analysis of the corresponding results of the vibrationresistance and the longitudinal wave propagation velocity yields the relation that the vibrationresistance is proportional to ( ) .c 1 24 . Although the vibration resistance is related to the wavepropagation velocity by a power equation, the curve plotted in Fig.6.12 for this relation isalmost a straight line. Hence, there is a certain linear relation between the vibrationresistance and the wave propagation velocity. Actually, the ratio of the vibration resistancein terms of ppv to the wave propagation velocity is the dynamic tensile strain capacity of theconcrete. Regression analysis of the dynamic tensile strain capacity results indicates that thevariation of the dynamic tensile strain capacity with the compressive strength of the concreteis very small and the dynamic tensile strain capacity may therefore be taken to have a constantvalue. The ratio of the vibration resistance to the wave propagation velocity is thus more orless constant, or in other words, the vibration resistance is approximately proportional to thewave propagation velocity. In view of the above regression analysis results, it is postulatedthat the longitudinal wave propagation velocity and the dynamic tensile strain capacity are themajor material parameters determining the vibration resistance of a concrete. Among thesetwo parameters, the dynamic tensile strain capacity of the concrete is of particular importancebecause the concrete fails when the dynamic tensile strain capacity is exceeded by thedynamic tensile strain but the dynamic tensile strain produced by a shock wave needs not bedependent on the longitudinal wave propagation velocity of the concrete. In an unrestrainedconcrete structure, the dynamic tensile strain induced is equal to the ppv of the vibrationdivided by the wave propagation velocity of the concrete, but in a restrained concretestructure which is forced to deform together with the surrounding soil, the strain induced inthe concrete will be affected by that induced in the soil which is equal to the ppv of thevibration divided by the wave propagation velocity of the soil. Further discussions on thistopic are given in the next section.

7.4 Restrained and Unrestrained Structures

Restrained and unrestrained structures respond quite differently to blasting vibrationsand should therefore be treated separately.

Restrained structures (mainly underground structures) are restrained to deform togetherwith the surrounding soil and thus the strain produced in the structure is affected by soil-structure interaction. Let the wave propagation velocity in the concrete be denoted by cand that in the soil by cs . If the concrete structure and the surrounding soil are free todeform independently, then the maximum strains εd and ε s produced in the concrete and inthe soil are given respectively by:

εd =vc

…(7.1)

- 150 -

εs =vcs

…(7.2)

However, due to interaction between the structure and the soil, the strains produced in themare forced to be equal and the actual strain produced in the composite system will besomewhere between the two values given by the above two equations. It is not easy toevaluate the effects of the soil-structure interaction. If the structure is relatively small suchthat the surrounding soil has a much larger volume than the structure, then the structure willbe forced to have its strain equal to that given by Eqn.7.2. On the other hand, if the structureis big, then the effects of the soil-structure interaction will be relatively small and the structurewill have its strain close to that given by Eqn.7.1. Any real situation lies between these twoextremes and a conservative estimate for the strain produced in the concrete structure may betaken as the larger of the two values given by Eqns. 7.1 and 7.2, as depicted below:

εd = max. ,vcvcs

=vc csmin.( , )

…(7.3)

When c cs < , the strain produced in the concrete structure will be determined by the value ofcs , or in other words, the ground strain induced by the blasting vibration may cause failure ofthe concrete structure. When c cs > , the strain produced in the concrete structure will bedetermined by the value of c , or in other words, the soil-structure interaction will not haveany detrimental effect on the structure.

Unrestrained structures (mainly aboveground structures) are unrestrained by anynearby soil and are therefore free to deform and vibrate. For an unrestrained structure, thestrain induced by the propagation of a shock wave through it is governed solely by Eqn.7.1and the ground strain induced by the shock wave will not affect the structure. However,since the structure is free to vibrate, the shock induced structural vibration needs to beconsidered as the structure will shake as if it is subjected to an earthquake. The shakingresponse of the structure to the blasting vibration depends on many factors including thenatural frequency and damping of the structure, and the intensity, duration and frequencyspectrum of the ground motion. Since the ground motions due to blasting are similar innature to those due to earthquakes, the dynamic response of a structure subjected to groundshaking due to blasting may be analyzed as in the seismic case.

The vibration resistance of concrete obtained in the present study and the vibrationcontrol limits derived therefrom are applicable only to unrestrained structures which are bulkyor massive such that they are unlikely to have any dynamic response to the blasting vibrationapart from the propagation of shock wave through them. When the vibration control limitsare applied to restrained structures, the effects of the ground strain induced in the surroundingsoil need also to be considered. In order to avoid possible damages due to large groundstrain, the ground strain should be limited to within the safe dynamic tensile strain capacity ofthe concrete. Hence, for restrained structures, the vibration control should be exercised suchthat both the following conditions are complied with:

(a) the ppv of the blasting vibration must not exceed the safe ppv limit of the concrete;

- 151 -

(b) the ground strain, evaluated as the ratio of the ppv of the blasting vibration to the wavepropagation velocity of the surrounding soil, must not exceed the safe dynamic tensilestrain capacity of the concrete.

For hardened concrete, i.e. concrete at age more than 12 hours, the condition (a) wouldbe more critical when c cs< while the condition (b) would be more critical when c cs> .For concrete at age less than 12 hours, the wave propagation velocity had not been measuredbut judging from its low stiffness, it may be assumed that c cs< and thus the condition (a)would govern. In fact, the dynamic tensile strain capacity of concrete at such early age hadnever been measured and the condition (b) cannot be directly applied. Nevertheless, forcompleteness, it is suggested to keep the condition (b) for concrete at age less than 12 hourswith the safe dynamic tensile strain capacity of the concrete taken as that of hardened concretewhich has a constant value of 32 micro-strain. It is considered a good practice to alwaysimpose condition (b) so that the ground strain would not cause damage to any nearbyunknown buried concrete structure. Condition (a) is a lot more restrictive than condition (b)when applied to concrete at age less than 12 hours and thus the enforcement of condition (b)would not change the vibration control limits for concrete at such age.

When the vibration control limits are applied to unrestrained structures which mayhave significant dynamic response to the ground motion caused by blasting, the stressesinduced by the dynamic response need also to be considered. In such case, detailed dynamicanalysis is required and the vibration control should be exercised such that both the followingconditions are complied with:

(a) the ppv of the blasting vibration must not exceed the safe ppv limit of the concrete;(b) the dynamic response of the structure to the ground motion due to blasting must not be

excessive as to cause damage to the structure.

In addition to the above, it should be advisable to always monitor the actual vibrationresponse of the structure and carry out regular inspections to see if any damage has beencaused. If the structure is already occupied, than the possibility of causing nuisance to thetenants and/or disturbance to any sensitive equipment housed inside should also be considered.Avoidance of nuisance and disturbance to people and sensitive equipment may impose a verysevere restriction to the blasting operation.

- 152 -

8. CONCLUSIONS AND RECOMMENDATIONS FOR FURTHER STUDY

8.1 Conclusions

An experimental programme of subjecting typical concrete used in Hong Kong to highintensity shock vibrations to evaluate their vibration resistances at different ages andinvestigate the short/long term effects of shock vibration on concrete is completed. Usingthe test results so gathered, a theoretical study to correlate the vibration resistance of aconcrete to its mechanical properties and develop a method of setting up appropriate vibrationcontrol criteria based on the known properties of concrete has been carried out. Theconclusions of the experimental investigation and the theoretical study are presented in thefollowings.

Method of testing:

(1) A method of testing fresh/hardened concrete for their resistance against shockvibration by subjecting prisms cast of the concrete to hammer blows has beendeveloped. The hammering operation is made automatic and synchronized with thedata logging process by the use of a computer controlled hammer release device. Apiezoelectric high-g accelerometer fixed at mid-length of the prism is used to pick upthe acceleration response of the concrete. Analogue output from the conditioningamplifier of the accelerometer is digitized by a computer controlled A/D converter andthen stored in a hard disk for subsequent digital signal processing. The velocityresponse of the concrete is obtained by numerical integration of the acceleration dataand then a trend removal process is applied to extract the pure wave component of thevelocity response from which the peak particle velocity of the shock vibration appliedis determined. The short term effects of the shock vibration applied are studied byobserving the occurrence of cracks on the surface of the concrete prism afterhammering while the long term effects are evaluated by continuing to cure theconcrete prism until the 28-day and then testing the prism for its tensile andcompressive strengths. The tensile strength of the concrete is measured by a directtension test method instead of the easier three-point beam flexural test method in orderto eliminate the unpredictable effects of the random positioning of the transversecracks formed by hammering on the test results. The broken pieces of the concreteprism after the tension test are then used to determine the compressive strength of theconcrete. This hammering test method has been shown to be a practical and efficientmethod of evaluating the short/long term effects of high-intensity shock vibration onconcrete. It does not require the use of explosives and can therefore be safely carriedout in a laboratory. Unlike previous test methods developed by others which do notbother to measure the tensile strength of concrete, the present method gives the effectsof the shock vibration on the tensile strength as well as the compressive strength. Infact, it is believed this is the first time the effects of shock vibration on the tensilestrength of concrete are studied in detail.

Material properties:

(2) As part of the experimental programme, a large amount of test results on the materialproperties of typical concrete used in Hong Kong has been produced. The concretesstudied are of grades 20 - 40, without PFA or with 25% PFA, and cured at 20oC or

- 154 -

avoid any damage to the concrete due to blasting and not to rely on a false hope ofself-healing.

Vibration resistance of concrete:

(5) Since it is not possible to continuously adjust the shock vibration intensity until theconcrete fails, the vibration resistance of concrete cannot be directly measured.Nevertheless, for this particular study in which shock vibrations of different intensitywere applied in such a way that some of the test prisms were cracked while the othersremained un-cracked, the vibration resistances of the concretes tested may beindirectly determined from the lowest intensity of shock vibration applied that hadcaused failure of the concrete (ppv1) and the highest intensity of shock vibrationapplied that had not caused any failure of the concrete (ppv2). The failure criterionadopted for determining the vibration resistance is “complete cracking or having morethan 30% reduction in tensile or compressive strength”. However, due to the erraticnature of the effects of shock vibration on concrete, ppv1 is not always greater thanppv2. From each pair of ppv1 and ppv2 values, a lower bound estimate of thevibration resistance (ppv3) may be obtained as min.(ppv1,ppv2). The ppv3 results ofthe 12 batches of concrete tested are listed in Tables 5.25 - 5.27.

(6) No obvious effects of using PFA up to 25% and curing temperature within the rangeof 20oC - 30oC on the vibration resistance of concrete can be identified from the ppv3results. Any probable effects of PFA and curing temperature within the rangesstudied have been over-shadowed by the large intrinsic variation of the vibrationresistance of concrete. In view of this situation, it is suggested that when settingvibration control limits to avoid vibration damages to concrete, a single set ofvibration limits that applies to concrete with or without PFA and cured within thetemperature range of 20oC - 30oC should be used; it does not seem to be worthwhile tohave more refined vibration control limits that take these two factors into account.

(7) Comparison of the lower bound estimates of the vibration resistances of the differentgrades of concrete in Figs. 5.1 and 5.2 reveals that no effect of concrete grade can bemeaningfully differentiated from the test results, although it is believed that theconcrete grade should have a bearing on the vibration resistance. To overcome thisdifficulty, a lower bound for all the concrete mixes tested, i.e. concretes of grade 20 -40, with 0 - 25 % PFA and cured at 20oC - 30oC, is worked out and used to establish asingle set of vibration control limits for them. The lower bound values of vibrationresistance are:

- 155 -

Age of concrete Lower bound of vibration resistance (mm/s)2 hrs. 464 hrs. 526 hrs. 1308 hrs. 18010 hrs. 17012 hrs. 25018 hrs. 18024 hrs. 3503 days 5007 days 69028 days 750

Correlation of vibration resistance to other mechanical properties:

(8) For theoretical study, mean value estimates of the vibration resistance (ppv4), whichare taken as equal to (ppv1+ppv2)/2 and should be more accurate than the lowerbound estimates, are used. For concrete of age less than 12 hours, it is found that thecorrelation between ppv4 and the penetration resistance of the mortar fraction of theconcrete is rather weak and that the vibration resistance increases only marginally withthe penetration resistance. For concrete of age more than 12 hours, correlation ofppv4 to the cube compressive strength f cu and split cylinder tensile strength f tyields the following equations:

ppv4 = 585 ⋅ ( ) .f cu0 26

ppv4 = 1098 ⋅ ( ) .f t0 29

in which the unit(s) for ppv4 is mm/s, and for compressive and tensile strengths areMPa. From these two equations, it can be seen that the vibration resistance is neitherproportional to the compressive strength nor to the tensile strength. The believe ofmany engineers that the vibration resistance is proportional to the strength of theconcrete is unfounded. Correlation of ppv4 to the longitudinal wave propagationvelocity c yields:

ppv4 = 0.047 ⋅ ( ) .c 1 24

in which the unit for wave propagation velocity is m/s. Correlation of the dynamictensile strain capacity εd ' , which is actually just the ratio of the vibration resistance interms of ppv to the longitudinal wave propagation velocity, gives the followingequation:

εd ' = 321 ⋅ ( )f cu0.015

where the unit for tensile strain capacity is micro-strain. It is seen from this equationthat the dynamic tensile strain capacity changes only slightly with the compressivestrength of the concrete and may therefore be taken to have a constant value. Hence,the ratio of the vibration resistance to the wave propagation velocity is more or lessconstant. In other words, the vibration resistance is approximately proportional to

- 156 -

the wave propagation velocity. It is thus postulated that the longitudinal wavepropagation velocity and the dynamic tensile strain capacity are the major materialparameters determining the vibration resistance of a concrete.

(9) However, the above mean value estimates are not suitable for direct use to establishvibration control limits. As there may be large intrinsic variation in vibrationresistance, the actual vibration resistance of the concrete could be significantly lowerthan the values predicted by the above equations. What is more important is thatthere is a very high probability of about 50% of having the actual vibration resistancelower than the mean value. Like the use of characteristic strength in structural design,it is proposed to use a characteristic value of the vibration resistance to cater for thelarge intrinsic variation of vibration resistance. The characteristic vibrationresistance (ppvc) is defined as the vibration level at which no more than 5% of theconcrete is expected to fail. Assuming that ppvc varies with the other mechanicalproperties of concrete in the same way as for ppv4 and applying curve fitting to thetest results, the following equations are obtained:

ppvc = 290 ⋅ ( ) .f cu0 26

ppvc = 527 ⋅ ( ) .f t0 29

ppvc = 0.023 ⋅ ( ) .c 1 24

The corresponding characteristic value of dynamic tensile strain capacity so derived isapproximately equal to 160 micro-strain.

(10) The measured vibration resistance is generally higher than the theoretical predictiongiven by the following equation:

vcEf

dtd' =

where v' is the theoretical vibration resistance in terms of ppv and f td is the dynamictensile strength which may be taken as 1.5 times the static tensile strength. In fact,the measured vibration resistance is often two to three times higher than the abovetheoretical value. Nevertheless, since some of the test results are just marginallyhigher than v' , the theoretical value given by this equation may be taken as a lowerbound to the vibration resistance.

Recommended vibration control limits:

(11) For the purpose of establishing vibration control limits, it is suggested to adopt arelatively large factor of safety of 5 to allow for the large intrinsic variation of thevibration resistance. A set of vibration control limits that applies to concretes ofgrade 20 - 40, with 0 - 25 % PFA and cured at 20oC - 30oC may be established fromthe lower bound values of vibration resistance given in Conclusion (7) by dividing thelower bound values by a factor of safety of 5 and adjusting the resulting values slightlyso that they map into a smooth and monotonic function of the age of concrete, aspresented below:

- 157 -

Age of concrete Safe ppv limit (mm/s)≤ 4 hrs. 10

6 - 8 hrs. 2010 - 12 hrs. 30

18 hrs. 4024 hrs. 703 days 1007 days 125

≥ 28 days 150

(12) Alternatively, the vibration control limits may also be set up based on the knownproperties of concrete using the equations for characteristic vibration resistance givenin Conclusion (9). Using the equation correlating the characteristic vibrationresistance to the cube compressive strength and dividing the ppvc values so obtainedby a factor of safety of 5, the following vibration control limits are established:

Cube compressive strength (MPa) Safe ppv limit (mm/s)3.0 775.0 8810.0 10620.0 12630.0 14040.0 15150.0 160

Using the equation correlating the characteristic vibration resistance to the splitcylinder tensile strength and dividing the ppvc values so obtained by a factor of safetyof 5, the following vibration control limits are established:

Split cylinder tensile strength (MPa) Safe ppv limit (mm/s)0.1 540.3 740.5 861.0 1052.0 1293.0 1454.0 158

(13) Since the vibration resistance is roughly proportional to the longitudinal wavepropagation velocity and the characteristic value of the dynamic tensile strain capacity(the dynamic tensile strain capacity is just the ratio of vibration resistance tolongitudinal wave propagation velocity) is approximately equal to a constant value of160 micro-strain, the vibration control limits may also be established by first dividingthe characteristic dynamic tensile strain capacity by a factor of safety of 5 to obtain thesafe dynamic tensile strain capacity as 32 micro-strain and then multiplying by thelongitudinal wave propagation velocity to produce the vibration control limits asfollows:

- 158 -

safe ppv limit= 32 micro-strain × longitudinal wave propagation velocity

(14) The above vibration control limits are applicable only to unrestrained structures whichare unlikely to have significant dynamic response to the blasting vibration apart fromthe propagation of shock wave through them. When the vibration control limits areapplied to restrained structures, the effects of the ground strain induced in thesurrounding soil need also to be considered. In order to avoid damages due to largeground strain, the ground strain should be limited to within the safe dynamic tensilestrain capacity of the concrete. Hence, for restrained concrete structures, thevibration control should be exercised such that both the following conditions arecomplied with:

(a) the ppv of the blasting vibration must not exceed the safe ppv limit of theconcrete;

(b) the ground strain, evaluated as the ratio of the ppv of the blasting vibration to thewave propagation velocity of the surrounding soil, must not exceed 32 micro-strain.

(15) When the vibration control limits are applied to unrestrained concrete structures whichmay have significant dynamic response to the ground motion caused by blasting, thestresses induced by the dynamic response need also to be considered. In such case,the vibration control should be exercised such that both the following conditions arecomplied with:

(a) the ppv of the blasting vibration must not exceed the safe ppv limit of theconcrete;

(b) the dynamic response of the structure to the ground motion due to blasting mustnot be excessive as to cause damage to the structure.

8.2 Recommendations for Further Study

The experimental investigation carried out in this project has been confined tolaboratory testing. So far, no field testing has been performed to measure the resistance ofconcrete to real blasting vibrations. It is, therefore, recommended to cast concrete on typicalconstruction sites where blasting is going to take place at different times before the blastingand at different distances from the blast to investigate the short/long term effects of blastingvibrations on concrete at various ages. However, it should be noted that especially for greenconcrete, the effects of the blasting vibrations may not be easily observable by naked eyes andthus the vibrated concrete, after being continuously cured until a certain age, should be takenout for testing of their tensile and compressive strengths. Since soil-structure interaction canaffect the response of the concrete, sites with different soil conditions should be selected forthe study. As the moulds may have some effects on the interaction between the concrete andthe soil, some of the concrete should be cast without moulds, i.e. directly against the soil,while some other concrete cast inside moulds so that any possible difference due to the use ofmoulds may be investigated.

- 159 -

9. REFERENCES

Bryson, B. and Cooley, T. (1985). “Blasting procedures, Veterans Administration MedicalCentre, Birmingham, Alabama”, Proceedings, 11th Conference on Explosives andBlasting Technique, Society of Explosive Engineers, pp.272-283.

Dowding, C.H. (1985). Blasting Vibration Monitoring and Control, Prentice Hall Inc., NewJersey, pp.184-186.

Gamble, D.L. and Simpson, T.A. (1985). “Effects of blasting vibrations on uncured concretefoundations”, Proceedings, 11th Conference on Explosives and Blasting Technique,Society of Explosive Engineers, pp.124-136.

Gardner, N.J. (1990). “Effect of temperature on the early-age properties of type I, type III, andtype I/fly ash concretes”, A.C.I. Materials Journal, Vol. 87, No.1, pp.68-78.

Hulshizer, A.J. and Desai, A.J. (1984). “Shock vibration effects on freshly placed concrete”,Journal of Construction Engineering and Management, A.S.C.E., Vol. 110, No.2,pp.266-285.

Kakizaki, M. et al. (1992). Effect of Mixing Method on Mechanical Properties and PoreStructure of Ultra-High Strength Concrete, Katri Report No.90, 19pp.

Lydon, F.D. and Balendran, R.V. (1986). “Some observations on elastic properties of plainconcrete”, Cement and Concrete Research, Vol.16, No.3, pp.314-324.

Neville, A.M. (1995). Properties of Concrete, Fourth Edition, Longman Group Limited,England, Chapter 6 and Chapter 9, pp269-317 and pp.412-481.

Olofesson, S.O. (1988). Applied Explosives Technology for Construction and Mining, Applex,pp.236-237.

Raphael, J.M. (1984). “Tensile strength of concrete”, A.C.I. Materials Journal, Vol.81, No.2,pp.158-165.

Wiss, J.F. (1981). “Construction vibrations: state-of-the-art”, Journal of the GeotechnicalEngineering Division, A.S.C.E., Vol. 107, No. GT2, February, pp.167-181.

a study of the effect of blasting vibration on green concrete

Documents